feat: simplify STTService request handling by refining SDK usage and improving error logging

🌊 feat: refine SDK usage logic in STT and TTS services, improve header handling
🌊 feat: enhance TTSService with Deepgram SDK integration and refactor voice validation
2024-12-04 17:33:38 +01:00 · 2024-11-24 01:14:17 +01:00 · 2024-11-23 16:49:56 +01:00 · 2024-11-23 12:20:17 +01:00 · 2024-11-23 12:17:53 +01:00
2091 changed files with 82770 additions and 226869 deletions
--- a/.env.example
+++ b/.env.example
@@ -15,37 +15,11 @@ HOST=localhost
 PORT=3080

 MONGO_URI=mongodb://127.0.0.1:27017/LibreChat
-#The maximum number of connections in the connection pool. */
-MONGO_MAX_POOL_SIZE=
-#The minimum number of connections in the connection pool. */
-MONGO_MIN_POOL_SIZE=
-#The maximum number of connections that may be in the process of being established concurrently by the connection pool. */
-MONGO_MAX_CONNECTING=
-#The maximum number of milliseconds that a connection can remain idle in the pool before being removed and closed. */
-MONGO_MAX_IDLE_TIME_MS=
-#The maximum time in milliseconds that a thread can wait for a connection to become available. */
-MONGO_WAIT_QUEUE_TIMEOUT_MS=
-# Set to false to disable automatic index creation for all models associated with this connection. */
-MONGO_AUTO_INDEX=
-# Set to `false` to disable Mongoose automatically calling `createCollection()` on every model created on this connection. */
-MONGO_AUTO_CREATE=

 DOMAIN_CLIENT=http://localhost:3080
 DOMAIN_SERVER=http://localhost:3080

 NO_INDEX=true
-# Use the address that is at most n number of hops away from the Express application.
-# req.socket.remoteAddress is the first hop, and the rest are looked for in the X-Forwarded-For header from right to left.
-# A value of 0 means that the first untrusted address would be req.socket.remoteAddress, i.e. there is no reverse proxy.
-# Defaulted to 1.
-TRUST_PROXY=1
-
-# Minimum password length for user authentication
-# Default: 8
-# Note: When using LDAP authentication, you may want to set this to 1 
-# to bypass local password validation, as LDAP servers handle their own
-# password policies.
-# MIN_PASSWORD_LENGTH=8

 #===============#
 # JSON Logging  #
@@ -79,7 +53,7 @@ DEBUG_CONSOLE=false
 #                     Endpoints                     #
 #===================================================#

-# ENDPOINTS=openAI,assistants,azureOpenAI,google,anthropic
+# ENDPOINTS=openAI,assistants,azureOpenAI,bingAI,google,gptPlugins,anthropic

 PROXY=

@@ -109,7 +83,7 @@ PROXY=
 #============#

 ANTHROPIC_API_KEY=user_provided
-# ANTHROPIC_MODELS=claude-opus-4-20250514,claude-sonnet-4-20250514,claude-3-7-sonnet-20250219,claude-3-5-sonnet-20241022,claude-3-5-haiku-20241022,claude-3-opus-20240229,claude-3-sonnet-20240229,claude-3-haiku-20240307
+# ANTHROPIC_MODELS=claude-3-5-haiku-20241022,claude-3-5-sonnet-20241022,claude-3-5-sonnet-latest,claude-3-5-sonnet-20240620,claude-3-opus-20240229,claude-3-sonnet-20240229,claude-3-haiku-20240307,claude-2.1,claude-2,claude-1.2,claude-1,claude-1-100k,claude-instant-1,claude-instant-1-100k
 # ANTHROPIC_REVERSE_PROXY=

 #============#
@@ -131,6 +105,13 @@ ANTHROPIC_API_KEY=user_provided
 # AZURE_OPENAI_API_EMBEDDINGS_DEPLOYMENT_NAME= # Deprecated
 # PLUGINS_USE_AZURE="true" # Deprecated

+#============#
+# BingAI     #
+#============#
+
+BINGAI_TOKEN=user_provided
+# BINGAI_HOST=https://cn.bing.com
+
 #=================#
 #   AWS Bedrock   #
 #=================#
@@ -138,7 +119,6 @@ ANTHROPIC_API_KEY=user_provided
 # BEDROCK_AWS_DEFAULT_REGION=us-east-1 # A default region must be provided
 # BEDROCK_AWS_ACCESS_KEY_ID=someAccessKey
 # BEDROCK_AWS_SECRET_ACCESS_KEY=someSecretAccessKey
-# BEDROCK_AWS_SESSION_TOKEN=someSessionToken

 # Note: This example list is not meant to be exhaustive. If omitted, all known, supported model IDs will be included for you.
 # BEDROCK_AWS_MODELS=anthropic.claude-3-5-sonnet-20240620-v1:0,meta.llama3-1-8b-instruct-v1:0
@@ -157,18 +137,15 @@ ANTHROPIC_API_KEY=user_provided
 #============#

 GOOGLE_KEY=user_provided
-
 # GOOGLE_REVERSE_PROXY=
-# Some reverse proxies do not support the X-goog-api-key header, uncomment to pass the API key in Authorization header instead.
-# GOOGLE_AUTH_HEADER=true

 # Gemini API (AI Studio)
-# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash,gemini-2.0-flash-lite
+# GOOGLE_MODELS=gemini-1.5-flash-latest,gemini-1.0-pro,gemini-1.0-pro-001,gemini-1.0-pro-latest,gemini-1.0-pro-vision-latest,gemini-1.5-pro-latest,gemini-pro,gemini-pro-vision

 # Vertex AI
-# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash-001,gemini-2.0-flash-lite-001
+# GOOGLE_MODELS=gemini-1.5-flash-preview-0514,gemini-1.5-pro-preview-0514,gemini-1.0-pro-vision-001,gemini-1.0-pro-002,gemini-1.0-pro-001,gemini-pro-vision,gemini-1.0-pro

-# GOOGLE_TITLE_MODEL=gemini-2.0-flash-lite-001
+# GOOGLE_TITLE_MODEL=gemini-pro

 # GOOGLE_LOC=us-central1

@@ -189,22 +166,21 @@ GOOGLE_KEY=user_provided
 # GOOGLE_SAFETY_HATE_SPEECH=BLOCK_ONLY_HIGH
 # GOOGLE_SAFETY_HARASSMENT=BLOCK_ONLY_HIGH
 # GOOGLE_SAFETY_DANGEROUS_CONTENT=BLOCK_ONLY_HIGH
-# GOOGLE_SAFETY_CIVIC_INTEGRITY=BLOCK_ONLY_HIGH

 #============#
 # OpenAI     #
 #============#

 OPENAI_API_KEY=user_provided
-# OPENAI_MODELS=o1,o1-mini,o1-preview,gpt-4o,gpt-4.5-preview,chatgpt-4o-latest,gpt-4o-mini,gpt-3.5-turbo-0125,gpt-3.5-turbo-0301,gpt-3.5-turbo,gpt-4,gpt-4-0613,gpt-4-vision-preview,gpt-3.5-turbo-0613,gpt-3.5-turbo-16k-0613,gpt-4-0125-preview,gpt-4-turbo-preview,gpt-4-1106-preview,gpt-3.5-turbo-1106,gpt-3.5-turbo-instruct,gpt-3.5-turbo-instruct-0914,gpt-3.5-turbo-16k
+# OPENAI_MODELS=gpt-4o,chatgpt-4o-latest,gpt-4o-mini,gpt-3.5-turbo-0125,gpt-3.5-turbo-0301,gpt-3.5-turbo,gpt-4,gpt-4-0613,gpt-4-vision-preview,gpt-3.5-turbo-0613,gpt-3.5-turbo-16k-0613,gpt-4-0125-preview,gpt-4-turbo-preview,gpt-4-1106-preview,gpt-3.5-turbo-1106,gpt-3.5-turbo-instruct,gpt-3.5-turbo-instruct-0914,gpt-3.5-turbo-16k

 DEBUG_OPENAI=false

 # TITLE_CONVO=false
-# OPENAI_TITLE_MODEL=gpt-4o-mini
+# OPENAI_TITLE_MODEL=gpt-3.5-turbo

 # OPENAI_SUMMARIZE=true
-# OPENAI_SUMMARY_MODEL=gpt-4o-mini
+# OPENAI_SUMMARY_MODEL=gpt-3.5-turbo

 # OPENAI_FORCE_PROMPT=true

@@ -230,6 +206,12 @@ ASSISTANTS_API_KEY=user_provided
 # More info, including how to enable use of Assistants with Azure here:
 # https://www.librechat.ai/docs/configuration/librechat_yaml/ai_endpoints/azure#using-assistants-with-azure

+#============#
+# OpenRouter #
+#============#
+# !!!Warning: Use the variable above instead of this one. Using this one will override the OpenAI endpoint
+# OPENROUTER_API_KEY=
+
 #============#
 # Plugins    #
 #============#
@@ -252,14 +234,6 @@ AZURE_AI_SEARCH_SEARCH_OPTION_QUERY_TYPE=
 AZURE_AI_SEARCH_SEARCH_OPTION_TOP=
 AZURE_AI_SEARCH_SEARCH_OPTION_SELECT=

-# OpenAI Image Tools Customization
-#----------------
-# IMAGE_GEN_OAI_DESCRIPTION_WITH_FILES=Custom description for image generation tool when files are present
-# IMAGE_GEN_OAI_DESCRIPTION_NO_FILES=Custom description for image generation tool when no files are present
-# IMAGE_EDIT_OAI_DESCRIPTION=Custom description for image editing tool
-# IMAGE_GEN_OAI_PROMPT_DESCRIPTION=Custom prompt description for image generation tool
-# IMAGE_EDIT_OAI_PROMPT_DESCRIPTION=Custom prompt description for image editing tool
-
 # DALL·E
 #----------------
 # DALLE_API_KEY=
@@ -277,23 +251,11 @@ AZURE_AI_SEARCH_SEARCH_OPTION_SELECT=
 # DALLE3_AZURE_API_VERSION=
 # DALLE2_AZURE_API_VERSION=

-# Flux
-#-----------------
-FLUX_API_BASE_URL=https://api.us1.bfl.ai
-# FLUX_API_BASE_URL = 'https://api.bfl.ml';
-
-# Get your API key at https://api.us1.bfl.ai/auth/profile
-# FLUX_API_KEY=
-
 # Google
 #-----------------
 GOOGLE_SEARCH_API_KEY=
 GOOGLE_CSE_ID=

-# YOUTUBE
-#-----------------
-YOUTUBE_API_KEY=
-
 # SerpAPI
 #-----------------
 SERPAPI_API_KEY=
@@ -327,10 +289,6 @@ MEILI_NO_ANALYTICS=true
 MEILI_HOST=http://0.0.0.0:7700
 MEILI_MASTER_KEY=DrhYf7zENyR6AlUCKmnz0eYASOQdl6zxH7s7MKFSfFCt

-# Optional: Disable indexing, useful in a multi-node setup
-# where only one instance should perform an index sync.
-# MEILI_NO_SYNC=true
-
 #==================================================#
 #          Speech to Text & Text to Speech         #
 #==================================================#
@@ -370,11 +328,6 @@ REGISTRATION_VIOLATION_SCORE=1
 CONCURRENT_VIOLATION_SCORE=1
 MESSAGE_VIOLATION_SCORE=1
 NON_BROWSER_VIOLATION_SCORE=20
-TTS_VIOLATION_SCORE=0
-STT_VIOLATION_SCORE=0
-FORK_VIOLATION_SCORE=0
-IMPORT_VIOLATION_SCORE=0
-FILE_UPLOAD_VIOLATION_SCORE=0

 LOGIN_MAX=7
 LOGIN_WINDOW=5
@@ -398,7 +351,7 @@ ILLEGAL_MODEL_REQ_SCORE=5
 # Balance                #
 #========================#

-# CHECK_BALANCE=false
+CHECK_BALANCE=false
 # START_BALANCE=20000 # note: the number of tokens that will be credited after registration.

 #========================#
@@ -433,22 +386,12 @@ FACEBOOK_CALLBACK_URL=/oauth/facebook/callback
 GITHUB_CLIENT_ID=
 GITHUB_CLIENT_SECRET=
 GITHUB_CALLBACK_URL=/oauth/github/callback
-# GitHub Enterprise
-# GITHUB_ENTERPRISE_BASE_URL=
-# GITHUB_ENTERPRISE_USER_AGENT=

 # Google
 GOOGLE_CLIENT_ID=
 GOOGLE_CLIENT_SECRET=
 GOOGLE_CALLBACK_URL=/oauth/google/callback

-# Apple
-APPLE_CLIENT_ID=
-APPLE_TEAM_ID=
-APPLE_KEY_ID=
-APPLE_PRIVATE_KEY_PATH=
-APPLE_CALLBACK_URL=/oauth/apple/callback
-
 # OpenID
 OPENID_CLIENT_ID=
 OPENID_CLIENT_SECRET=
@@ -463,95 +406,18 @@ OPENID_REQUIRED_ROLE_PARAMETER_PATH=
 OPENID_USERNAME_CLAIM=
 # Set to determine which user info property returned from OpenID Provider to store as the User's name
 OPENID_NAME_CLAIM=
-# Optional audience parameter for OpenID authorization requests
-OPENID_AUDIENCE=

 OPENID_BUTTON_LABEL=
 OPENID_IMAGE_URL=
-# Set to true to automatically redirect to the OpenID provider when a user visits the login page
-# This will bypass the login form completely for users, only use this if OpenID is your only authentication method
-OPENID_AUTO_REDIRECT=false
-# Set to true to use PKCE (Proof Key for Code Exchange) for OpenID authentication
-OPENID_USE_PKCE=false
-#Set to true to reuse openid tokens for authentication management instead of using the mongodb session and the custom refresh token.
-OPENID_REUSE_TOKENS=
-#By default, signing key verification results are cached in order to prevent excessive HTTP requests to the JWKS endpoint.
-#If a signing key matching the kid is found, this will be cached and the next time this kid is requested the signing key will be served from the cache.
-#Default is true.
-OPENID_JWKS_URL_CACHE_ENABLED=
-OPENID_JWKS_URL_CACHE_TIME= # 600000 ms eq to 10 minutes leave empty to disable caching
-#Set to true to trigger token exchange flow to acquire access token for the userinfo endpoint.
-OPENID_ON_BEHALF_FLOW_FOR_USERINFO_REQUIRED=
-OPENID_ON_BEHALF_FLOW_USERINFO_SCOPE="user.read" # example for Scope Needed for Microsoft Graph API
-# Set to true to use the OpenID Connect end session endpoint for logout
-OPENID_USE_END_SESSION_ENDPOINT=
-
-#========================#
-# SharePoint Integration #
-#========================#
-# Requires Entra ID (OpenID) authentication to be configured
-
-# Enable SharePoint file picker in chat and agent panels
-# ENABLE_SHAREPOINT_FILEPICKER=true
-
-# SharePoint tenant base URL (e.g., https://yourtenant.sharepoint.com)
-# SHAREPOINT_BASE_URL=https://yourtenant.sharepoint.com
-
-# Microsoft Graph API And SharePoint scopes for file picker
-# SHAREPOINT_PICKER_SHAREPOINT_SCOPE==https://yourtenant.sharepoint.com/AllSites.Read
-# SHAREPOINT_PICKER_GRAPH_SCOPE=Files.Read.All
-#========================#
-
-# SAML
-# Note: If OpenID is enabled, SAML authentication will be automatically disabled.
-SAML_ENTRY_POINT=
-SAML_ISSUER=
-SAML_CERT=
-SAML_CALLBACK_URL=/oauth/saml/callback
-SAML_SESSION_SECRET=
-
-# Attribute mappings (optional)
-SAML_EMAIL_CLAIM=
-SAML_USERNAME_CLAIM=
-SAML_GIVEN_NAME_CLAIM=
-SAML_FAMILY_NAME_CLAIM=
-SAML_PICTURE_CLAIM=
-SAML_NAME_CLAIM=
-
-# Logint buttion settings (optional)
-SAML_BUTTON_LABEL=
-SAML_IMAGE_URL=
-
-# Whether the SAML Response should be signed.
-# - If "true", the entire `SAML Response` will be signed.
-# - If "false" or unset, only the `SAML Assertion` will be signed (default behavior).
-# SAML_USE_AUTHN_RESPONSE_SIGNED=
-
-
-#===============================================#
-# Microsoft Graph API / Entra ID Integration  #
-#===============================================#
-
-# Enable Entra ID people search integration in permissions/sharing system
-# When enabled, the people picker will search both local database and Entra ID
-USE_ENTRA_ID_FOR_PEOPLE_SEARCH=false
-
-# When enabled, entra id groups owners will be considered as members of the group
-ENTRA_ID_INCLUDE_OWNERS_AS_MEMBERS=false
-
-# Microsoft Graph API scopes needed for people/group search
-# Default scopes provide access to user profiles and group memberships
-OPENID_GRAPH_SCOPES=User.Read,People.Read,GroupMember.Read.All

 # LDAP
 LDAP_URL=
 LDAP_BIND_DN=
 LDAP_BIND_CREDENTIALS=
 LDAP_USER_SEARCH_BASE=
-#LDAP_SEARCH_FILTER="mail="
+LDAP_SEARCH_FILTER=mail={{username}}
 LDAP_CA_CERT_PATH=
 # LDAP_TLS_REJECT_UNAUTHORIZED=
-# LDAP_STARTTLS=
 # LDAP_LOGIN_USES_USERNAME=true
 # LDAP_ID=
 # LDAP_USERNAME=
@@ -573,18 +439,6 @@ EMAIL_PASSWORD=
 EMAIL_FROM_NAME=
 EMAIL_FROM=noreply@librechat.ai

-#========================#
-#      Mailgun API       #
-#========================#
-
-# MAILGUN_API_KEY=your-mailgun-api-key
-# MAILGUN_DOMAIN=mg.yourdomain.com
-# EMAIL_FROM=noreply@yourdomain.com
-# EMAIL_FROM_NAME="LibreChat"
-
-# # Optional: For EU region
-# MAILGUN_HOST=https://api.eu.mailgun.net
-
 #========================#
 # Firebase CDN           #
 #========================#
@@ -596,24 +450,6 @@ FIREBASE_STORAGE_BUCKET=
 FIREBASE_MESSAGING_SENDER_ID=
 FIREBASE_APP_ID=

-#========================#
-# S3 AWS Bucket          #
-#========================#
-
-AWS_ENDPOINT_URL=
-AWS_ACCESS_KEY_ID=
-AWS_SECRET_ACCESS_KEY=
-AWS_REGION=
-AWS_BUCKET_NAME=
-
-#========================#
-# Azure Blob Storage     #
-#========================#
-
-AZURE_STORAGE_CONNECTION_STRING=
-AZURE_STORAGE_PUBLIC_ACCESS=false
-AZURE_CONTAINER_NAME=files
-
 #========================#
 # Shared Links           #
 #========================#
@@ -633,10 +469,6 @@ ALLOW_SHARED_LINKS_PUBLIC=true
 # If you have another service in front of your LibreChat doing compression, disable express based compression here
 # DISABLE_COMPRESSION=true

-# If you have gzipped version of uploaded image images in the same folder, this will enable gzip scan and serving of these images
-# Note: The images folder will be scanned on startup and a ma kept in memory. Be careful for large number of images.
-# ENABLE_IMAGE_OUTPUT_GZIP_SCAN=true
-
 #===================================================#
 #                        UI                         #
 #===================================================#
@@ -650,49 +482,6 @@ HELP_AND_FAQ_URL=https://librechat.ai
 # Google tag manager id
 #ANALYTICS_GTM_ID=user provided google tag manager id

-#===============#
-# REDIS Options #
-#===============#
-
-# Enable Redis for caching and session storage
-# USE_REDIS=true
-
-# Single Redis instance
-# REDIS_URI=redis://127.0.0.1:6379
-
-# Redis cluster (multiple nodes)
-# REDIS_URI=redis://127.0.0.1:7001,redis://127.0.0.1:7002,redis://127.0.0.1:7003
-
-# Redis with TLS/SSL encryption and CA certificate
-# REDIS_URI=rediss://127.0.0.1:6380
-# REDIS_CA=/path/to/ca-cert.pem
-
-# Elasticache may need to use an alternate dnsLookup for TLS connections.  see "Special Note: Aws Elasticache Clusters with TLS" on this webpage: https://www.npmjs.com/package/ioredis
-# Enable alternative dnsLookup for redis
-# REDIS_USE_ALTERNATIVE_DNS_LOOKUP=true
-
-# Redis authentication (if required)
-# REDIS_USERNAME=your_redis_username
-# REDIS_PASSWORD=your_redis_password
-
-# Redis key prefix configuration
-# Use environment variable name for dynamic prefix (recommended for cloud deployments)
-# REDIS_KEY_PREFIX_VAR=K_REVISION
-# Or use static prefix directly
-# REDIS_KEY_PREFIX=librechat
-
-# Redis connection limits
-# REDIS_MAX_LISTENERS=40
-
-# Redis ping interval in seconds (0 = disabled, >0 = enabled)
-# When set to a positive integer, Redis clients will ping the server at this interval to keep connections alive
-# When unset or 0, no pinging is performed (recommended for most use cases)
-# REDIS_PING_INTERVAL=300
-
-# Force specific cache namespaces to use in-memory storage even when Redis is enabled
-# Comma-separated list of CacheKeys (e.g., ROLES,MESSAGES)
-# FORCED_IN_MEMORY_CACHE_NAMESPACES=ROLES,MESSAGES
-
 #==================================================#
 #                      Others                      #
 #==================================================#
@@ -700,6 +489,9 @@ HELP_AND_FAQ_URL=https://librechat.ai

 # NODE_ENV=

+# REDIS_URI=
+# USE_REDIS=
+
 # E2E_USER_EMAIL=
 # E2E_USER_PASSWORD=

@@ -711,58 +503,10 @@ HELP_AND_FAQ_URL=https://librechat.ai
 #   users always get the latest version. Customize    #
 #   only if you understand caching implications.      #

-# INDEX_CACHE_CONTROL=no-cache, no-store, must-revalidate
-# INDEX_PRAGMA=no-cache
-# INDEX_EXPIRES=0
+# INDEX_HTML_CACHE_CONTROL=no-cache, no-store, must-revalidate
+# INDEX_HTML_PRAGMA=no-cache
+# INDEX_HTML_EXPIRES=0

 # no-cache: Forces validation with server before using cached version
 # no-store: Prevents storing the response entirely
-# must-revalidate: Prevents using stale content when offline
-
-#=====================================================#
-#                  OpenWeather                        #
-#=====================================================#
-OPENWEATHER_API_KEY=
-
-#====================================#
-# LibreChat Code Interpreter API     #
-#====================================#
-
-# https://code.librechat.ai
-# LIBRECHAT_CODE_API_KEY=your-key
-
-#======================#
-# Web Search           #
-#======================#
-
-# Note: All of the following variable names can be customized.
-# Omit values to allow user to provide them.
-
-# For more information on configuration values, see:
-# https://librechat.ai/docs/features/web_search
-
-# Search Provider (Required)
-# SERPER_API_KEY=your_serper_api_key
-
-# Scraper (Required)
-# FIRECRAWL_API_KEY=your_firecrawl_api_key
-# Optional: Custom Firecrawl API URL
-# FIRECRAWL_API_URL=your_firecrawl_api_url
-
-# Reranker (Required)
-# JINA_API_KEY=your_jina_api_key
-# or
-# COHERE_API_KEY=your_cohere_api_key
-
-#======================#
-# MCP Configuration    #
-#======================#
-
-# Treat 401/403 responses as OAuth requirement when no oauth metadata found
-# MCP_OAUTH_ON_AUTH_ERROR=true
-
-# Timeout for OAuth detection requests in milliseconds
-# MCP_OAUTH_DETECTION_TIMEOUT=5000
-
-# Cache connection status checks for this many milliseconds to avoid expensive verification
-# MCP_CONNECTION_CHECK_TTL=60000
+# must-revalidate: Prevents using stale content when offline
--- a/.eslintrc.js
+++ b/.eslintrc.js
@@ -0,0 +1,173 @@
+module.exports = {
+  env: {
+    browser: true,
+    es2021: true,
+    node: true,
+    commonjs: true,
+    es6: true,
+  },
+  extends: [
+    'eslint:recommended',
+    'plugin:react/recommended',
+    'plugin:react-hooks/recommended',
+    'plugin:jest/recommended',
+    'prettier',
+    'plugin:jsx-a11y/recommended',
+  ],
+  ignorePatterns: [
+    'client/dist/**/*',
+    'client/public/**/*',
+    'e2e/playwright-report/**/*',
+    'packages/data-provider/types/**/*',
+    'packages/data-provider/dist/**/*',
+    'packages/data-provider/test_bundle/**/*',
+    'data-node/**/*',
+    'meili_data/**/*',
+    'node_modules/**/*',
+  ],
+  parser: '@typescript-eslint/parser',
+  parserOptions: {
+    ecmaVersion: 'latest',
+    sourceType: 'module',
+    ecmaFeatures: {
+      jsx: true,
+    },
+  },
+  plugins: ['react', 'react-hooks', '@typescript-eslint', 'import', 'jsx-a11y'],
+  rules: {
+    'react/react-in-jsx-scope': 'off',
+    '@typescript-eslint/ban-ts-comment': ['error', { 'ts-ignore': 'allow' }],
+    indent: ['error', 2, { SwitchCase: 1 }],
+    'max-len': [
+      'error',
+      {
+        code: 120,
+        ignoreStrings: true,
+        ignoreTemplateLiterals: true,
+        ignoreComments: true,
+      },
+    ],
+    'linebreak-style': 0,
+    curly: ['error', 'all'],
+    semi: ['error', 'always'],
+    'object-curly-spacing': ['error', 'always'],
+    'no-multiple-empty-lines': ['error', { max: 1 }],
+    'no-trailing-spaces': 'error',
+    'comma-dangle': ['error', 'always-multiline'],
+    // "arrow-parens": [2, "as-needed", { requireForBlockBody: true }],
+    // 'no-plusplus': ['error', { allowForLoopAfterthoughts: true }],
+    'no-console': 'off',
+    'import/no-cycle': 'error',
+    'import/no-self-import': 'error',
+    'import/extensions': 'off',
+    'no-promise-executor-return': 'off',
+    'no-param-reassign': 'off',
+    'no-continue': 'off',
+    'no-restricted-syntax': 'off',
+    'react/prop-types': ['off'],
+    'react/display-name': ['off'],
+    'no-nested-ternary': 'error',
+    'no-unused-vars': ['error', { varsIgnorePattern: '^_' }],
+    quotes: ['error', 'single'],
+  },
+  overrides: [
+    {
+      files: ['**/*.ts', '**/*.tsx'],
+      rules: {
+        'no-unused-vars': 'off', // off because it conflicts with '@typescript-eslint/no-unused-vars'
+        'react/display-name': 'off',
+        '@typescript-eslint/no-unused-vars': 'warn',
+      },
+    },
+    {
+      files: ['rollup.config.js', '.eslintrc.js', 'jest.config.js'],
+      env: {
+        node: true,
+      },
+    },
+    {
+      files: [
+        '**/*.test.js',
+        '**/*.test.jsx',
+        '**/*.test.ts',
+        '**/*.test.tsx',
+        '**/*.spec.js',
+        '**/*.spec.jsx',
+        '**/*.spec.ts',
+        '**/*.spec.tsx',
+        'setupTests.js',
+      ],
+      env: {
+        jest: true,
+        node: true,
+      },
+      rules: {
+        'react/display-name': 'off',
+        'react/prop-types': 'off',
+        'react/no-unescaped-entities': 'off',
+      },
+    },
+    {
+      files: ['**/*.ts', '**/*.tsx'],
+      parser: '@typescript-eslint/parser',
+      parserOptions: {
+        project: './client/tsconfig.json',
+      },
+      plugins: ['@typescript-eslint/eslint-plugin', 'jest'],
+      extends: [
+        'plugin:@typescript-eslint/eslint-recommended',
+        'plugin:@typescript-eslint/recommended',
+      ],
+      rules: {
+        '@typescript-eslint/no-explicit-any': 'error',
+        '@typescript-eslint/no-unnecessary-condition': 'warn',
+        '@typescript-eslint/strict-boolean-expressions': 'warn',
+      },
+    },
+    {
+      files: './packages/data-provider/**/*.ts',
+      overrides: [
+        {
+          files: '**/*.ts',
+          parser: '@typescript-eslint/parser',
+          parserOptions: {
+            project: './packages/data-provider/tsconfig.json',
+          },
+        },
+      ],
+    },
+    {
+      files: './config/translations/**/*.ts',
+      parser: '@typescript-eslint/parser',
+      parserOptions: {
+        project: './config/translations/tsconfig.json',
+      },
+    },
+    {
+      files: ['./packages/data-provider/specs/**/*.ts'],
+      parserOptions: {
+        project: './packages/data-provider/tsconfig.spec.json',
+      },
+    },
+  ],
+  settings: {
+    react: {
+      createClass: 'createReactClass', // Regex for Component Factory to use,
+      // default to "createReactClass"
+      pragma: 'React', // Pragma to use, default to "React"
+      fragment: 'Fragment', // Fragment to use (may be a property of <pragma>), default to "Fragment"
+      version: 'detect', // React version. "detect" automatically picks the version you have installed.
+    },
+    'import/parsers': {
+      '@typescript-eslint/parser': ['.ts', '.tsx'],
+    },
+    'import/resolver': {
+      typescript: {
+        project: ['./client/tsconfig.json'],
+      },
+      node: {
+        project: ['./client/tsconfig.json'],
+      },
+    },
+  },
+};
--- a/.github/CONTRIBUTING.md
+++ b/.github/CONTRIBUTING.md
@@ -24,40 +24,22 @@ Project maintainers have the right and responsibility to remove, edit, or reject

 ## To contribute to this project, please adhere to the following guidelines:

-## 1. Development Setup
+## 1. Development notes

-1. Use Node.JS 20.x.
-2. Install typescript globally: `npm i -g typescript`.
-3. Run `npm ci` to install dependencies.
-4. Build the data provider: `npm run build:data-provider`.
-5. Build data schemas: `npm run build:data-schemas`.
-6. Build API methods: `npm run build:api`.
-7. Setup and run unit tests:
-    - Copy `.env.test`: `cp api/test/.env.test.example api/test/.env.test`.
-    - Run backend unit tests: `npm run test:api`.
-    - Run frontend unit tests: `npm run test:client`.
-8. Setup and run integration tests:
-    - Build client: `cd client && npm run build`.
-    - Create `.env`: `cp .env.example .env`.
-    - Install [MongoDB Community Edition](https://www.mongodb.com/docs/manual/administration/install-community/), ensure that `mongosh` connects to your local instance.
-    - Run: `npx install playwright`, then `npx playwright install`.
-    - Copy `config.local`: `cp e2e/config.local.example.ts e2e/config.local.ts`.
-    - Copy `librechat.yaml`: `cp librechat.example.yaml librechat.yaml`.
-    - Run: `npm run e2e`.
-
-## 2. Development Notes
-
-1. Before starting work, make sure your main branch has the latest commits with `npm run update`.
-3. Run linting command to find errors: `npm run lint`. Alternatively, ensure husky pre-commit checks are functioning.
+1. Before starting work, make sure your main branch has the latest commits with `npm run update`
+2. Run linting command to find errors: `npm run lint`. Alternatively, ensure husky pre-commit checks are functioning.
 3. After your changes, reinstall packages in your current branch using `npm run reinstall` and ensure everything still works. 
    - Restart the ESLint server ("ESLint: Restart ESLint Server" in VS Code command bar) and your IDE after reinstalling or updating.
 4. Clear web app localStorage and cookies before and after changes.
-5. For frontend changes, compile typescript before and after changes to check for introduced errors: `cd client && npm run build`.
-6. Run backend unit tests: `npm run test:api`.
-7. Run frontend unit tests: `npm run test:client`.
-8. Run integration tests: `npm run e2e`.
+5. For frontend changes:
+    - Install typescript globally: `npm i -g typescript`.
+    - Compile typescript before and after changes to check for introduced errors: `cd client && tsc --noEmit`.
+6. Run tests locally:
+    - Backend unit tests: `npm run test:api`
+    - Frontend unit tests: `npm run test:client`
+    - Integration tests: `npm run e2e` (requires playwright installed, `npx install playwright`)

-## 3. Git Workflow
+## 2. Git Workflow

 We utilize a GitFlow workflow to manage changes to this project's codebase. Follow these general steps when contributing code:

@@ -67,7 +49,7 @@ We utilize a GitFlow workflow to manage changes to this project's codebase. Foll
 4. Submit a pull request with a clear and concise description of your changes and the reasons behind them.
 5. We will review your pull request, provide feedback as needed, and eventually merge the approved changes into the main branch.

-## 4. Commit Message Format
+## 3. Commit Message Format

 We follow the [semantic format](https://gist.github.com/joshbuchea/6f47e86d2510bce28f8e7f42ae84c716) for commit messages.

@@ -94,7 +76,7 @@ feat: add hat wobble
 ```


-## 5. Pull Request Process
+## 4. Pull Request Process

 When submitting a pull request, please follow these guidelines:

@@ -109,7 +91,7 @@ Ensure that your changes meet the following criteria:
 - The commit history is clean and easy to follow. You can use `git rebase` or `git merge --squash` to clean your commit history before submitting the pull request.
 - The pull request description clearly outlines the changes and the reasons behind them. Be sure to include the steps to test the pull request.

-## 6. Naming Conventions
+## 5. Naming Conventions

 Apply the following naming conventions to branches, labels, and other Git-related entities:

@@ -118,7 +100,7 @@ Apply the following naming conventions to branches, labels, and other Git-relate
 - **JS/TS:** Directories and file names: Descriptive and camelCase. First letter uppercased for React files (e.g., `helperFunction.ts, ReactComponent.tsx`).
 - **Docs:** Directories and file names: Descriptive and snake_case (e.g., `config_files.md`).

-## 7. TypeScript Conversion
+## 6. TypeScript Conversion

 1. **Original State**: The project was initially developed entirely in JavaScript (JS).

@@ -144,10 +126,10 @@ Apply the following naming conventions to branches, labels, and other Git-relate
      
   - **Current Stance**: At present, this backend transition is of lower priority and might not be pursued.

-## 8. Module Import Conventions
+## 7. Module Import Conventions

 - `npm` packages first, 
-     - from longest line (top) to shortest (bottom)
+     - from shortest line (top) to longest (bottom)

 - Followed by typescript types (pertains to data-provider and client workspaces)
     - longest line (top) to shortest (bottom)
@@ -157,8 +139,6 @@ Apply the following naming conventions to branches, labels, and other Git-relate
     - longest line (top) to shortest (bottom)
     - imports with alias `~` treated the same as relative import with respect to line length

-**Note:** ESLint will automatically enforce these import conventions when you run `npm run lint --fix` or through pre-commit hooks.
-
 ---

 Please ensure that you adapt this summary to fit the specific context and nuances of your project.
--- a/.github/ISSUE_TEMPLATE/BUG-REPORT.yml
+++ b/.github/ISSUE_TEMPLATE/BUG-REPORT.yml
@@ -1,19 +1,12 @@
 name: Bug Report
 description: File a bug report
 title: "[Bug]: "
-labels: ["🐛 bug"]
+labels: ["bug"]
 body:
  - type: markdown
    attributes:
      value: |
        Thanks for taking the time to fill out this bug report!
-        
-        Before submitting, please:
-        - Search existing [Issues and Discussions](https://github.com/danny-avila/LibreChat/discussions) to see if your bug has already been reported
-        - Use [Discussions](https://github.com/danny-avila/LibreChat/discussions) instead of Issues for:
-          - General inquiries
-          - Help with setup
-          - Questions about whether you're experiencing a bug
  - type: textarea
    id: what-happened
    attributes:
@@ -22,23 +15,6 @@ body:
      placeholder: Please give as many details as possible
    validations:
      required: true
-  - type: textarea
-    id: version-info
-    attributes:
-      label: Version Information
-      description: |
-        If using Docker, please run and provide the output of:
-        ```bash
-        docker images | grep librechat
-        ```
-        
-        If running from source, please run and provide the output of:
-        ```bash
-        git rev-parse HEAD
-        ```
-      placeholder: Paste the output here
-    validations:
-      required: true
  - type: textarea
    id: steps-to-reproduce
    attributes:
@@ -63,24 +39,8 @@ body:
    id: logs
    attributes:
      label: Relevant log output
-      description: |
-        Please paste relevant logs that were created when reproducing the error.
-        
-        Log locations:
-        - Docker: Project root directory ./logs
-        - npm: ./api/logs
-        
-        There are two types of logs that can help diagnose the issue:
-        - debug logs (debug-YYYY-MM-DD.log)
-        - error logs (error-YYYY-MM-DD.log)
-        
-        Error logs contain exact stack traces and are especially helpful, but both can provide valuable information.
-        Please only include the relevant portions of logs that correspond to when you reproduced the error.
-
-        For UI-related issues, browser console logs can be very helpful. You can provide these as screenshots or paste the text here.
+      description: Please copy and paste any relevant log output. This will be automatically formatted into code, so no need for backticks.
      render: shell
-    validations:
-      required: true
  - type: textarea
    id: screenshots
    attributes:
@@ -93,4 +53,4 @@ body:
      description: By submitting this issue, you agree to follow our [Code of Conduct](https://github.com/danny-avila/LibreChat/blob/main/.github/CODE_OF_CONDUCT.md)
      options:
        - label: I agree to follow this project's Code of Conduct
-          required: true
+          required: true
--- a/.github/ISSUE_TEMPLATE/FEATURE-REQUEST.yml
+++ b/.github/ISSUE_TEMPLATE/FEATURE-REQUEST.yml
@@ -1,7 +1,7 @@
 name: Feature Request
 description: File a feature request
-title: "[Enhancement]: "
-labels: ["✨ enhancement"]
+title: "Enhancement: "
+labels: ["enhancement"]
 body:
  - type: markdown
    attributes:
--- a/.github/ISSUE_TEMPLATE/LOCIZE_TRANSLATION_ACCESS_REQUEST.yml
+++ b/.github/ISSUE_TEMPLATE/LOCIZE_TRANSLATION_ACCESS_REQUEST.yml
@@ -1,42 +0,0 @@
-name: Locize Translation Access Request
-description: Request access to an additional language in Locize for LibreChat translations.
-title: "Locize Access Request: "
-labels: ["🌍 i18n", "🔑 access request"]
-body:
-  - type: markdown
-    attributes:
-      value: |
-        Thank you for your interest in contributing to LibreChat translations!  
-        Please fill out the form below to request access to an additional language in **Locize**.
-
-        **🔗 Available Languages:** [View the list here](https://www.librechat.ai/docs/translation)
-        
-        **📌 Note:** Ensure that the requested language is supported before submitting your request.
-  - type: input
-    id: account_name
-    attributes:
-      label: Locize Account Name
-      description: Please provide your Locize account name (e.g., John Doe).
-      placeholder: e.g., John Doe
-    validations:
-      required: true
-  - type: input
-    id: language_requested
-    attributes:
-      label: Language Code (ISO 639-1)
-      description: |
-        Enter the **ISO 639-1** language code for the language you want to translate into.  
-        Example: `es` for Spanish, `zh-Hant` for Traditional Chinese.
-        
-        **🔗 Reference:** [Available Languages](https://www.librechat.ai/docs/translation)
-      placeholder: e.g., es
-    validations:
-      required: true
-  - type: checkboxes
-    id: agreement
-    attributes:
-      label: Agreement
-      description: By submitting this request, you confirm that you will contribute responsibly and adhere to the project guidelines.
-      options:
-        - label: I agree to use my access solely for contributing to LibreChat translations.
-          required: true
--- a/.github/ISSUE_TEMPLATE/NEW-LANGUAGE-REQUEST.yml
+++ b/.github/ISSUE_TEMPLATE/NEW-LANGUAGE-REQUEST.yml
@@ -1,33 +0,0 @@
-name: New Language Request
-description: Request to add a new language for LibreChat translations.
-title: "New Language Request: "
-labels: ["✨ enhancement", "🌍 i18n"]
-body:
-  - type: markdown
-    attributes:
-      value: |
-        Thank you for taking the time to submit a new language request! Please fill out the following details so we can review your request.
-  - type: input
-    id: language_name
-    attributes:
-      label: Language Name
-      description: Please provide the full name of the language (e.g., Spanish, Mandarin).
-      placeholder: e.g., Spanish
-    validations:
-      required: true
-  - type: input
-    id: iso_code
-    attributes:
-      label: ISO 639-1 Code
-      description: Please provide the ISO 639-1 code for the language (e.g., es for Spanish). You can refer to [this list](https://www.w3schools.com/tags/ref_language_codes.asp) for valid codes.
-      placeholder: e.g., es
-    validations:
-      required: true
-  - type: checkboxes
-    id: terms
-    attributes:
-      label: Code of Conduct
-      description: By submitting this issue, you agree to follow our [Code of Conduct](https://github.com/danny-avila/LibreChat/blob/main/.github/CODE_OF_CONDUCT.md).
-      options:
-        - label: I agree to follow this project's Code of Conduct
-          required: true
--- a/.github/ISSUE_TEMPLATE/QUESTION.yml
+++ b/.github/ISSUE_TEMPLATE/QUESTION.yml
@@ -0,0 +1,50 @@
+name: Question
+description: Ask your question
+title: "[Question]: "
+labels: ["question"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for taking the time to fill this!
+  - type: textarea
+    id: what-is-your-question
+    attributes:
+      label: What is your question?
+      description: Please give as many details as possible
+      placeholder: Please give as many details as possible
+    validations:
+      required: true
+  - type: textarea
+    id: more-details
+    attributes:
+      label: More Details
+      description: Please provide more details if needed.
+      placeholder: Please provide more details if needed.
+    validations:
+      required: true
+  - type: dropdown
+    id: browsers
+    attributes:
+      label: What is the main subject of your question?
+      multiple: true
+      options:
+        - Documentation
+        - Installation
+        - UI
+        - Endpoints
+        - User System/OAuth
+        - Other
+  - type: textarea
+    id: screenshots
+    attributes:
+      label: Screenshots
+      description: If applicable, add screenshots to help explain your problem. You can drag and drop, paste images directly here or link to them.
+  - type: checkboxes
+    id: terms
+    attributes:
+      label: Code of Conduct
+      description: By submitting this issue, you agree to follow our [Code of Conduct](https://github.com/danny-avila/LibreChat/blob/main/.github/CODE_OF_CONDUCT.md)
+      options:
+        - label: I agree to follow this project's Code of Conduct
+          required: true
--- a/.github/configuration-release.json
+++ b/.github/configuration-release.json
@@ -1,60 +0,0 @@
-{
-  "categories": [
-    {
-      "title": "### ✨ New Features",
-      "labels": ["feat"]
-    },
-    {
-      "title": "### 🌍 Internationalization",
-      "labels": ["i18n"]
-    },
-    {
-      "title": "### 👐 Accessibility",
-      "labels": ["a11y"]
-    },
-    {
-      "title": "### 🔧 Fixes",
-      "labels": ["Fix", "fix"]
-    },
-    {
-      "title": "### ⚙️ Other Changes",
-      "labels": ["ci", "style", "docs", "refactor", "chore"]
-    }
-  ],
-  "ignore_labels": [
-    "🔁 duplicate",
-    "📊 analytics",
-    "🌱 good first issue",
-    "🔍 investigation",
-    "🙏 help wanted",
-    "❌ invalid",
-    "❓ question",
-    "🚫 wontfix",
-    "🚀 release",
-    "version"
-  ],
-  "base_branches": ["main"],
-  "sort": {
-    "order": "ASC",
-    "on_property": "mergedAt"
-  },
-  "label_extractor": [
-    {
-      "pattern": "^(?:[^A-Za-z0-9]*)(feat|fix|chore|docs|refactor|ci|style|a11y|i18n)\\s*:",
-      "target": "$1",
-      "flags": "i",
-      "on_property": "title",
-      "method": "match"
-    },
-    {
-      "pattern": "^(?:[^A-Za-z0-9]*)(v\\d+\\.\\d+\\.\\d+(?:-rc\\d+)?).*",
-      "target": "version",
-      "flags": "i",
-      "on_property": "title",
-      "method": "match"
-    }
-  ],
-  "template": "## [#{{TO_TAG}}] - #{{TO_TAG_DATE}}\n\nChanges from #{{FROM_TAG}} to #{{TO_TAG}}.\n\n#{{CHANGELOG}}\n\n[See full release details][release-#{{TO_TAG}}]\n\n[release-#{{TO_TAG}}]: https://github.com/#{{OWNER}}/#{{REPO}}/releases/tag/#{{TO_TAG}}\n\n---",
-  "pr_template": "- #{{TITLE}} by **@#{{AUTHOR}}** in [##{{NUMBER}}](#{{URL}})",
-  "empty_template": "- no changes"
-}
--- a/.github/configuration-unreleased.json
+++ b/.github/configuration-unreleased.json
@@ -1,68 +0,0 @@
-{
-  "categories": [
-    {
-      "title": "### ✨ New Features",
-      "labels": ["feat"]
-    },
-    {
-      "title": "### 🌍 Internationalization",
-      "labels": ["i18n"]
-    },
-    {
-      "title": "### 👐 Accessibility",
-      "labels": ["a11y"]
-    },
-    {
-      "title": "### 🔧 Fixes",
-      "labels": ["Fix", "fix"]
-    },
-    {
-      "title": "### ⚙️ Other Changes",
-      "labels": ["ci", "style", "docs", "refactor", "chore"]
-    }
-  ],
-  "ignore_labels": [
-    "🔁 duplicate",
-    "📊 analytics",
-    "🌱 good first issue",
-    "🔍 investigation",
-    "🙏 help wanted",
-    "❌ invalid",
-    "❓ question",
-    "🚫 wontfix",
-    "🚀 release",
-    "version",
-    "action"
-  ],
-  "base_branches": ["main"],
-  "sort": {
-    "order": "ASC",
-    "on_property": "mergedAt"
-  },
-  "label_extractor": [
-    {
-      "pattern": "^(?:[^A-Za-z0-9]*)(feat|fix|chore|docs|refactor|ci|style|a11y|i18n)\\s*:",
-      "target": "$1",
-      "flags": "i",
-      "on_property": "title",
-      "method": "match"
-    },
-    {
-      "pattern": "^(?:[^A-Za-z0-9]*)(v\\d+\\.\\d+\\.\\d+(?:-rc\\d+)?).*",
-      "target": "version",
-      "flags": "i",
-      "on_property": "title",
-      "method": "match"
-    },
-    {
-      "pattern": "^(?:[^A-Za-z0-9]*)(action)\\b.*",
-      "target": "action",
-      "flags": "i",
-      "on_property": "title",
-      "method": "match"
-    }
-  ],
-  "template": "## [Unreleased]\n\n#{{CHANGELOG}}\n\n---",
-  "pr_template": "- #{{TITLE}} by **@#{{AUTHOR}}** in [##{{NUMBER}}](#{{URL}})",
-  "empty_template": "- no changes"
-}
--- a/.github/workflows/backend-review.yml
+++ b/.github/workflows/backend-review.yml
@@ -7,7 +7,6 @@ on:
      - release/*
    paths:
      - 'api/**'
-      - 'packages/**'
 jobs:
  tests_Backend:
    name: Run Backend unit tests
@@ -34,15 +33,9 @@ jobs:
      - name: Install dependencies
        run: npm ci

-      - name: Install Data Provider Package
+      - name: Install Data Provider
        run: npm run build:data-provider
-
-      - name: Install Data Schemas Package
-        run: npm run build:data-schemas
-
-      - name: Install API Package
-        run: npm run build:api
-
+      
      - name: Create empty auth.json file
        run: |
          mkdir -p api/data
@@ -67,8 +60,7 @@ jobs:
      - name: Run librechat-data-provider unit tests
        run: cd packages/data-provider && npm run test:ci

-      - name: Run @librechat/data-schemas unit tests
-        run: cd packages/data-schemas && npm run test:ci
-
-      - name: Run @librechat/api unit tests
-        run: cd packages/api && npm run test:ci
+      - name: Run linters
+        uses: wearerequired/lint-action@v2
+        with:
+          eslint: true
--- a/.github/workflows/client.yml
+++ b/.github/workflows/client.yml
@@ -1,58 +0,0 @@
-name: Publish `@librechat/client` to NPM
-
-on:
-  push:
-    branches:
-      - main
-    paths:
-      - 'packages/client/package.json'
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: false
-        default: 'Manual publish requested'
-
-jobs:
-  build-and-publish:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      
-      - name: Use Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '20.x'
-        
-      - name: Install client dependencies
-        run: cd packages/client && npm ci
-        
-      - name: Build client
-        run: cd packages/client && npm run build
-        
-      - name: Set up npm authentication
-        run: echo "//registry.npmjs.org/:_authToken=${{ secrets.PUBLISH_NPM_TOKEN }}" > ~/.npmrc
-        
-      - name: Check version change
-        id: check
-        working-directory: packages/client
-        run: |
-          PACKAGE_VERSION=$(node -p "require('./package.json').version")
-          PUBLISHED_VERSION=$(npm view @librechat/client version 2>/dev/null || echo "0.0.0")
-          if [ "$PACKAGE_VERSION" = "$PUBLISHED_VERSION" ]; then
-            echo "No version change, skipping publish"
-            echo "skip=true" >> $GITHUB_OUTPUT
-          else
-            echo "Version changed, proceeding with publish"
-            echo "skip=false" >> $GITHUB_OUTPUT
-          fi
-            
-      - name: Pack package
-        if: steps.check.outputs.skip != 'true'
-        working-directory: packages/client
-        run: npm pack
-        
-      - name: Publish
-        if: steps.check.outputs.skip != 'true'
-        working-directory: packages/client
-        run: npm publish *.tgz --access public
--- a/.github/workflows/data-provider.yml
+++ b/.github/workflows/data-provider.yml
@@ -1,4 +1,4 @@
-name: Publish `librechat-data-provider` to NPM
+name: Node.js Package

 on:
  push:
@@ -6,12 +6,6 @@ on:
      - main
    paths:
      - 'packages/data-provider/package.json'
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: false
-        default: 'Manual publish requested'

 jobs:
  build:
@@ -20,7 +14,7 @@ jobs:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
-          node-version: 20
+          node-version: 16
      - run: cd packages/data-provider && npm ci
      - run: cd packages/data-provider && npm run build

@@ -31,7 +25,7 @@ jobs:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
-          node-version: 20
+          node-version: 16
          registry-url: 'https://registry.npmjs.org'
      - run: cd packages/data-provider && npm ci
      - run: cd packages/data-provider && npm run build
--- a/.github/workflows/data-schemas.yml
+++ b/.github/workflows/data-schemas.yml
@@ -1,58 +0,0 @@
-name: Publish `@librechat/data-schemas` to NPM
-
-on:
-  push:
-    branches:
-      - main
-    paths:
-      - 'packages/data-schemas/package.json'
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: false
-        default: 'Manual publish requested'
-
-jobs:
-  build-and-publish:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      
-      - name: Use Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '20.x'
-          
-      - name: Install dependencies
-        run: cd packages/data-schemas && npm ci
-        
-      - name: Build
-        run: cd packages/data-schemas && npm run build
-        
-      - name: Set up npm authentication
-        run: echo "//registry.npmjs.org/:_authToken=${{ secrets.PUBLISH_NPM_TOKEN }}" > ~/.npmrc
-        
-      - name: Check version change
-        id: check
-        working-directory: packages/data-schemas
-        run: |
-          PACKAGE_VERSION=$(node -p "require('./package.json').version")
-          PUBLISHED_VERSION=$(npm view @librechat/data-schemas version 2>/dev/null || echo "0.0.0")
-          if [ "$PACKAGE_VERSION" = "$PUBLISHED_VERSION" ]; then
-            echo "No version change, skipping publish"
-            echo "skip=true" >> $GITHUB_OUTPUT
-          else
-            echo "Version changed, proceeding with publish"
-            echo "skip=false" >> $GITHUB_OUTPUT
-          fi
-            
-      - name: Pack package
-        if: steps.check.outputs.skip != 'true'
-        working-directory: packages/data-schemas
-        run: npm pack
-        
-      - name: Publish
-        if: steps.check.outputs.skip != 'true'
-        working-directory: packages/data-schemas
-        run: npm publish *.tgz --access public
--- a/.github/workflows/deploy-dev.yml
+++ b/.github/workflows/deploy-dev.yml
@@ -2,7 +2,7 @@ name: Update Test Server

 on:
  workflow_run:
-    workflows: ["Docker Dev Branch Images Build"]
+    workflows: ["Docker Dev Images Build"]
    types:
      - completed
  workflow_dispatch:
@@ -12,8 +12,7 @@ jobs:
    runs-on: ubuntu-latest
    if: |
      github.repository == 'danny-avila/LibreChat' &&
-      (github.event_name == 'workflow_dispatch' || 
-       (github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'dev'))
+      (github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success')
    steps:
    - name: Checkout repository
      uses: actions/checkout@v4
@@ -30,17 +29,13 @@ jobs:
        DO_USER: ${{ secrets.DO_USER }}
      run: |
        ssh -o StrictHostKeyChecking=no ${DO_USER}@${DO_HOST} << EOF
-        sudo -i -u danny bash << 'EEOF'
+        sudo -i -u danny bash << EEOF
        cd ~/LibreChat && \
        git fetch origin main && \
-        sudo npm run stop:deployed && \
-        sudo docker images --format "{{.Repository}}:{{.ID}}" | grep -E "lc-dev|librechat" | cut -d: -f2 | xargs -r sudo docker rmi -f || true && \
-        sudo npm run update:deployed && \
-        git checkout dev && \
-        git pull origin dev && \
+        npm run update:deployed && \
        git checkout do-deploy && \
-        git rebase dev && \
-        sudo npm run start:deployed && \
+        git rebase main && \
+        npm run start:deployed && \
        echo "Update completed. Application should be running now."
        EEOF
        EOF
--- a/.github/workflows/dev-branch-images.yml
+++ b/.github/workflows/dev-branch-images.yml
@@ -1,72 +0,0 @@
-name: Docker Dev Branch Images Build
-
-on:
-  workflow_dispatch:
-  push:
-    branches:
-      - dev
-    paths:
-      - 'api/**'
-      - 'client/**'
-      - 'packages/**'
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        include:
-          - target: api-build
-            file: Dockerfile.multi
-            image_name: lc-dev-api
-          - target: node
-            file: Dockerfile
-            image_name: lc-dev
-
-    steps:
-      # Check out the repository
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      # Set up QEMU
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
-
-      # Set up Docker Buildx
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      # Log in to GitHub Container Registry
-      - name: Log in to GitHub Container Registry
-        uses: docker/login-action@v2
-        with:
-          registry: ghcr.io
-          username: ${{ github.actor }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      # Login to Docker Hub
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_TOKEN }}
-
-      # Prepare the environment
-      - name: Prepare environment
-        run: |
-          cp .env.example .env
-
-      # Build and push Docker images for each target
-      - name: Build and push Docker images
-        uses: docker/build-push-action@v5
-        with:
-          context: .
-          file: ${{ matrix.file }}
-          push: true
-          tags: |
-            ghcr.io/${{ github.repository_owner }}/${{ matrix.image_name }}:${{ github.sha }}
-            ghcr.io/${{ github.repository_owner }}/${{ matrix.image_name }}:latest
-            ${{ secrets.DOCKERHUB_USERNAME }}/${{ matrix.image_name }}:${{ github.sha }}
-            ${{ secrets.DOCKERHUB_USERNAME }}/${{ matrix.image_name }}:latest
-          platforms: linux/amd64,linux/arm64
-          target: ${{ matrix.target }} 
--- a/.github/workflows/eslint-ci.yml
+++ b/.github/workflows/eslint-ci.yml
@@ -1,73 +0,0 @@
-name: ESLint Code Quality Checks
-
-on:
-  pull_request:
-    branches:
-      - main
-      - dev
-      - release/*
-    paths:
-      - 'api/**'
-      - 'client/**'
-
-jobs:
-  eslint_checks:
-    name: Run ESLint Linting
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      security-events: write
-      actions: read
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Set up Node.js 20.x
-        uses: actions/setup-node@v4
-        with:
-          node-version: 20
-          cache: npm
-
-      - name: Install dependencies
-        run: npm ci
-
-      # Run ESLint on changed files within the api/ and client/ directories.
-      - name: Run ESLint on changed files
-        env:
-          SARIF_ESLINT_IGNORE_SUPPRESSED: "true"
-        run: |
-          # Extract the base commit SHA from the pull_request event payload.
-          BASE_SHA=$(jq --raw-output .pull_request.base.sha "$GITHUB_EVENT_PATH")
-          echo "Base commit SHA: $BASE_SHA"
-
-          # Get changed files (only JS/TS files in api/ or client/)
-          CHANGED_FILES=$(git diff --name-only --diff-filter=ACMRTUXB "$BASE_SHA" HEAD | grep -E '^(api|client)/.*\.(js|jsx|ts|tsx)$' || true)
-
-          # Debug output
-          echo "Changed files:"
-          echo "$CHANGED_FILES"
-
-          # Ensure there are files to lint before running ESLint
-          if [[ -z "$CHANGED_FILES" ]]; then
-            echo "No matching files changed. Skipping ESLint."
-            echo "UPLOAD_SARIF=false" >> $GITHUB_ENV
-            exit 0
-          fi
-
-          # Set variable to allow SARIF upload
-          echo "UPLOAD_SARIF=true" >> $GITHUB_ENV
-
-          # Run ESLint
-          npx eslint --no-error-on-unmatched-pattern \
-            --config eslint.config.mjs \
-            --format @microsoft/eslint-formatter-sarif \
-            --output-file eslint-results.sarif $CHANGED_FILES || true
-
-      - name: Upload analysis results to GitHub
-        if: env.UPLOAD_SARIF == 'true'
-        uses: github/codeql-action/upload-sarif@v3
-        with:
-          sarif_file: eslint-results.sarif
-          wait-for-processing: true
--- a/.github/workflows/frontend-review.yml
+++ b/.github/workflows/frontend-review.yml
@@ -8,7 +8,7 @@ on:
      - release/*
    paths:
      - 'client/**'
-      - 'packages/data-provider/**'
+      - 'packages/**'

 jobs:
  tests_frontend_ubuntu:
--- a/.github/workflows/helmcharts.yml
+++ b/.github/workflows/helmcharts.yml
@@ -4,13 +4,12 @@ name: Build Helm Charts on Tag
 on:
  push:
    tags:
-      - "chart-*"
+      - "*"

 jobs:
  release:
    permissions:
      contents: write
-      packages: write
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
@@ -27,49 +26,8 @@ jobs:
        uses: azure/setup-helm@v4
        env:
          GITHUB_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
-      
-      - name: Build Subchart Deps
-        run: |
-          cd helm/librechat
-          helm dependency build  
-          cd ../librechat-rag-api
-          helm dependency build      

-      - name: Get Chart Version
-        id: chart-version
-        run: |
-          CHART_VERSION=$(echo "${{ github.ref_name }}" | cut -d'-' -f2)
-          echo "CHART_VERSION=${CHART_VERSION}" >> "$GITHUB_OUTPUT"
-
-      # Log in to GitHub Container Registry
-      - name: Log in to GitHub Container Registry
-        uses: docker/login-action@v3
-        with:
-          registry: ghcr.io
-          username: ${{ github.actor }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      # Run Helm OCI Charts Releaser
-      # This is for the librechat chart
-      - name: Release Helm OCI Charts for librechat
-        uses: appany/helm-oci-chart-releaser@v0.4.2
-        with:
-          name: librechat
-          repository: ${{ github.actor }}/librechat-chart
-          tag: ${{ steps.chart-version.outputs.CHART_VERSION }}
-          path: helm/librechat
-          registry: ghcr.io
-          registry_username: ${{ github.actor }}
-          registry_password: ${{ secrets.GITHUB_TOKEN }}
-      
-      # this is for the librechat-rag-api chart
-      - name: Release Helm OCI Charts for librechat-rag-api
-        uses: appany/helm-oci-chart-releaser@v0.4.2
-        with:
-          name: librechat-rag-api
-          repository: ${{ github.actor }}/librechat-chart
-          tag: ${{ steps.chart-version.outputs.CHART_VERSION }}
-          path: helm/librechat-rag-api
-          registry: ghcr.io
-          registry_username: ${{ github.actor }}
-          registry_password: ${{ secrets.GITHUB_TOKEN }}
+      - name: Run chart-releaser
+        uses: helm/chart-releaser-action@v1.6.0
+        env:
+          CR_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
--- a/.github/workflows/i18n-unused-keys.yml
+++ b/.github/workflows/i18n-unused-keys.yml
@@ -1,149 +0,0 @@
-name: Detect Unused i18next Strings
-
-# This workflow checks for unused i18n keys in translation files.
-# It has special handling for:
-# - com_ui_special_var_* keys that are dynamically constructed
-# - com_agents_category_* keys that are stored in the database and used dynamically
-
-on:
-  pull_request:
-    paths:
-      - "client/src/**"
-      - "api/**"
-      - "packages/data-provider/src/**"
-      - "packages/client/**"
-      - "packages/data-schemas/src/**"
-
-jobs:
-  detect-unused-i18n-keys:
-    runs-on: ubuntu-latest
-    permissions:
-      pull-requests: write
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v3
-
-      - name: Find unused i18next keys
-        id: find-unused
-        run: |
-          echo "🔍 Scanning for unused i18next keys..."
-
-          # Define paths
-          I18N_FILE="client/src/locales/en/translation.json"
-          SOURCE_DIRS=("client/src" "api" "packages/data-provider/src" "packages/client" "packages/data-schemas/src")
-
-          # Check if translation file exists
-          if [[ ! -f "$I18N_FILE" ]]; then
-            echo "::error title=Missing i18n File::Translation file not found: $I18N_FILE"
-            exit 1
-          fi
-
-          # Extract all keys from the JSON file
-          KEYS=$(jq -r 'keys[]' "$I18N_FILE")
-
-          # Track unused keys
-          UNUSED_KEYS=()
-
-          # Check if each key is used in the source code
-          for KEY in $KEYS; do
-            FOUND=false
-            
-            # Special case for dynamically constructed special variable keys
-            if [[ "$KEY" == com_ui_special_var_* ]]; then
-              # Check if TSpecialVarLabel is used in the codebase
-              for DIR in "${SOURCE_DIRS[@]}"; do
-                if grep -r --include=\*.{js,jsx,ts,tsx} -q "TSpecialVarLabel" "$DIR"; then
-                  FOUND=true
-                  break
-                fi
-              done
-              
-              # Also check if the key is directly used somewhere
-              if [[ "$FOUND" == false ]]; then
-                for DIR in "${SOURCE_DIRS[@]}"; do
-                  if grep -r --include=\*.{js,jsx,ts,tsx} -q "$KEY" "$DIR"; then
-                    FOUND=true
-                    break
-                  fi
-                done
-              fi
-            # Special case for agent category keys that are dynamically used from database
-            elif [[ "$KEY" == com_agents_category_* ]]; then
-              # Check if agent category localization is being used
-              for DIR in "${SOURCE_DIRS[@]}"; do
-                # Check for dynamic category label/description usage
-                if grep -r --include=\*.{js,jsx,ts,tsx} -E "category\.(label|description).*startsWith.*['\"]com_" "$DIR" > /dev/null 2>&1 || \
-                   # Check for the method that defines these keys
-                   grep -r --include=\*.{js,jsx,ts,tsx} "ensureDefaultCategories" "$DIR" > /dev/null 2>&1 || \
-                   # Check for direct usage in agentCategory.ts
-                   grep -r --include=\*.ts -E "label:.*['\"]$KEY['\"]" "$DIR" > /dev/null 2>&1 || \
-                   grep -r --include=\*.ts -E "description:.*['\"]$KEY['\"]" "$DIR" > /dev/null 2>&1; then
-                  FOUND=true
-                  break
-                fi
-              done
-              
-              # Also check if the key is directly used somewhere
-              if [[ "$FOUND" == false ]]; then
-                for DIR in "${SOURCE_DIRS[@]}"; do
-                  if grep -r --include=\*.{js,jsx,ts,tsx} -q "$KEY" "$DIR"; then
-                    FOUND=true
-                    break
-                  fi
-                done
-              fi
-            else
-              # Regular check for other keys
-              for DIR in "${SOURCE_DIRS[@]}"; do
-                if grep -r --include=\*.{js,jsx,ts,tsx} -q "$KEY" "$DIR"; then
-                  FOUND=true
-                  break
-                fi
-              done
-            fi
-          
-            if [[ "$FOUND" == false ]]; then
-              UNUSED_KEYS+=("$KEY")
-            fi
-          done
-
-          # Output results
-          if [[ ${#UNUSED_KEYS[@]} -gt 0 ]]; then
-            echo "🛑 Found ${#UNUSED_KEYS[@]} unused i18n keys:"
-            echo "unused_keys=$(echo "${UNUSED_KEYS[@]}" | jq -R -s -c 'split(" ")')" >> $GITHUB_ENV
-            for KEY in "${UNUSED_KEYS[@]}"; do
-              echo "::warning title=Unused i18n Key::'$KEY' is defined but not used in the codebase."
-            done
-          else
-            echo "✅ No unused i18n keys detected!"
-            echo "unused_keys=[]" >> $GITHUB_ENV
-          fi
-
-      - name: Post verified comment on PR
-        if: env.unused_keys != '[]'
-        run: |
-          PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH")
-
-          # Format the unused keys list as checkboxes for easy manual checking.
-          FILTERED_KEYS=$(echo "$unused_keys" | jq -r '.[]' | grep -v '^\s*$' | sed 's/^/- [ ] `/;s/$/`/' )
-
-          COMMENT_BODY=$(cat <<EOF
-          ### 🚨 Unused i18next Keys Detected
-
-          The following translation keys are defined in \`translation.json\` but are **not used** in the codebase:
-
-          $FILTERED_KEYS
-
-          ⚠️ **Please remove these unused keys to keep the translation files clean.**
-          EOF
-          )
-
-          gh api "repos/${{ github.repository }}/issues/${PR_NUMBER}/comments" \
-            -f body="$COMMENT_BODY" \
-            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}"
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Fail workflow if unused keys found
-        if: env.unused_keys != '[]'
-        run: exit 1
--- a/.github/workflows/locize-i18n-sync.yml
+++ b/.github/workflows/locize-i18n-sync.yml
@@ -1,72 +0,0 @@
-name: Sync Locize Translations & Create Translation PR
-
-on:
-  push:
-    branches: [main]
-  repository_dispatch:
-    types: [locize/versionPublished]
-
-jobs:
-  sync-translations:
-    name: Sync Translation Keys with Locize
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout Repository
-        uses: actions/checkout@v4
-
-      - name: Set Up Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: 20
-
-      - name: Install locize CLI
-        run: npm install -g locize-cli
-
-      # Sync translations (Push missing keys & remove deleted ones)
-      - name: Sync Locize with Repository
-        if: ${{ github.event_name == 'push' }}
-        run: |
-          cd client/src/locales
-          locize sync --api-key ${{ secrets.LOCIZE_API_KEY }} --project-id ${{ secrets.LOCIZE_PROJECT_ID }} --language en
-
-      # When triggered by repository_dispatch, skip sync step.
-      - name: Skip sync step on non-push events
-        if: ${{ github.event_name != 'push' }}
-        run: echo "Skipping sync as the event is not a push."
-
-  create-pull-request:
-    name: Create Translation PR on Version Published
-    runs-on: ubuntu-latest
-    needs: sync-translations
-    permissions:
-      contents: write
-      pull-requests: write
-    steps:
-      # 1. Check out the repository.
-      - name: Checkout Repository
-        uses: actions/checkout@v4
-
-      # 2. Download translation files from locize.
-      - name: Download Translations from locize
-        uses: locize/download@v2
-        with:
-          project-id: ${{ secrets.LOCIZE_PROJECT_ID }}
-          path: "client/src/locales"
-
-      # 3. Create a Pull Request using built-in functionality.
-      - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
-        with:
-          token: ${{ secrets.GITHUB_TOKEN }}
-          sign-commits: true
-          commit-message: "🌍 i18n: Update translation.json with latest translations"
-          base: main
-          branch: i18n/locize-translation-update
-          reviewers: danny-avila
-          title: "🌍 i18n: Update translation.json with latest translations"
-          body: |
-            **Description**:
-            - 🎯 **Objective**: Update `translation.json` with the latest translations from locize.
-            - 🔍 **Details**: This PR is automatically generated upon receiving a versionPublished event with version "latest". It reflects the newest translations provided by locize.
-            - ✅ **Status**: Ready for review.
-          labels: "🌍 i18n"
--- a/.github/workflows/unused-packages.yml
+++ b/.github/workflows/unused-packages.yml
@@ -1,244 +0,0 @@
-name: Detect Unused NPM Packages
-
-on:
-  pull_request:
-    paths:
-      - 'package.json'
-      - 'package-lock.json'
-      - 'client/**'
-      - 'api/**'
-      - 'packages/client/**'
-
-jobs:
-  detect-unused-packages:
-    runs-on: ubuntu-latest
-    permissions:
-      pull-requests: write
-
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: Use Node.js 20.x
-        uses: actions/setup-node@v4
-        with:
-          node-version: 20
-          cache: 'npm'
-
-      - name: Install depcheck
-        run: npm install -g depcheck
-
-      - name: Validate JSON files
-        run: |
-          for FILE in package.json client/package.json api/package.json packages/client/package.json; do
-            if [[ -f "$FILE" ]]; then
-              jq empty "$FILE" || (echo "::error title=Invalid JSON::$FILE is invalid" && exit 1)
-            fi
-          done
-
-      - name: Extract Dependencies Used in Scripts
-        id: extract-used-scripts
-        run: |
-          extract_deps_from_scripts() {
-            local package_file=$1
-            if [[ -f "$package_file" ]]; then
-              jq -r '.scripts | to_entries[].value' "$package_file" | \
-                grep -oE '([a-zA-Z0-9_-]+)' | sort -u > used_scripts.txt
-            else
-              touch used_scripts.txt
-            fi
-          }
-
-          extract_deps_from_scripts "package.json"
-          mv used_scripts.txt root_used_deps.txt
-
-          extract_deps_from_scripts "client/package.json"
-          mv used_scripts.txt client_used_deps.txt
-
-          extract_deps_from_scripts "api/package.json"
-          mv used_scripts.txt api_used_deps.txt
-
-      - name: Extract Dependencies Used in Source Code
-        id: extract-used-code
-        run: |
-          extract_deps_from_code() {
-            local folder=$1
-            local output_file=$2
-            if [[ -d "$folder" ]]; then
-              # Extract require() statements
-              grep -rEho "require\\(['\"]([a-zA-Z0-9@/._-]+)['\"]\\)" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
-                sed -E "s/require\\(['\"]([a-zA-Z0-9@/._-]+)['\"]\\)/\1/" > "$output_file"
-          
-              # Extract ES6 imports - various patterns
-              # import x from 'module'
-              grep -rEho "import .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
-                sed -E "s/import .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
-              
-              # import 'module' (side-effect imports)
-              grep -rEho "import ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
-                sed -E "s/import ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
-              
-              # export { x } from 'module' or export * from 'module'
-              grep -rEho "export .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
-                sed -E "s/export .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
-              
-              # import type { x } from 'module' (TypeScript)
-              grep -rEho "import type .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{ts,tsx} | \
-                sed -E "s/import type .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
-          
-              # Remove subpath imports but keep the base package
-              # e.g., '@tanstack/react-query/devtools' becomes '@tanstack/react-query'
-              sed -i -E 's|^(@?[a-zA-Z0-9-]+(/[a-zA-Z0-9-]+)?)/.*|\1|' "$output_file"
-              
-              sort -u "$output_file" -o "$output_file"
-            else
-              touch "$output_file"
-            fi
-          }
-
-          extract_deps_from_code "." root_used_code.txt
-          extract_deps_from_code "client" client_used_code.txt
-          extract_deps_from_code "api" api_used_code.txt
-          
-          # Extract dependencies used by @librechat/client package
-          extract_deps_from_code "packages/client" packages_client_used_code.txt
-
-      - name: Get @librechat/client dependencies
-        id: get-librechat-client-deps
-        run: |
-          if [[ -f "packages/client/package.json" ]]; then
-            # Get all dependencies from @librechat/client (dependencies, devDependencies, and peerDependencies)
-            DEPS=$(jq -r '.dependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
-            DEV_DEPS=$(jq -r '.devDependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
-            PEER_DEPS=$(jq -r '.peerDependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
-            
-            # Combine all dependencies
-            echo "$DEPS" > librechat_client_deps.txt
-            echo "$DEV_DEPS" >> librechat_client_deps.txt
-            echo "$PEER_DEPS" >> librechat_client_deps.txt
-            
-            # Also include dependencies that are imported in packages/client
-            cat packages_client_used_code.txt >> librechat_client_deps.txt
-            
-            # Remove empty lines and sort
-            grep -v '^$' librechat_client_deps.txt | sort -u > temp_deps.txt
-            mv temp_deps.txt librechat_client_deps.txt
-          else
-            touch librechat_client_deps.txt
-          fi
-
-      - name: Extract Workspace Dependencies
-        id: extract-workspace-deps
-        run: |
-          # Function to get dependencies from a workspace package that are used by another package
-          get_workspace_package_deps() {
-            local package_json=$1
-            local output_file=$2
-            
-            # Get all workspace dependencies (starting with @librechat/)
-            if [[ -f "$package_json" ]]; then
-              local workspace_deps=$(jq -r '.dependencies // {} | to_entries[] | select(.key | startswith("@librechat/")) | .key' "$package_json" 2>/dev/null || echo "")
-              
-              # For each workspace dependency, get its dependencies
-              for dep in $workspace_deps; do
-                # Convert @librechat/api to packages/api
-                local workspace_path=$(echo "$dep" | sed 's/@librechat\//packages\//')
-                local workspace_package_json="${workspace_path}/package.json"
-                
-                if [[ -f "$workspace_package_json" ]]; then
-                  # Extract all dependencies from the workspace package
-                  jq -r '.dependencies // {} | keys[]' "$workspace_package_json" 2>/dev/null >> "$output_file"
-                  # Also extract peerDependencies
-                  jq -r '.peerDependencies // {} | keys[]' "$workspace_package_json" 2>/dev/null >> "$output_file"
-                fi
-              done
-            fi
-            
-            if [[ -f "$output_file" ]]; then
-              sort -u "$output_file" -o "$output_file"
-            else
-              touch "$output_file"
-            fi
-          }
-          
-          # Get workspace dependencies for each package
-          get_workspace_package_deps "package.json" root_workspace_deps.txt
-          get_workspace_package_deps "client/package.json" client_workspace_deps.txt
-          get_workspace_package_deps "api/package.json" api_workspace_deps.txt
-
-      - name: Run depcheck for root package.json
-        id: check-root
-        run: |
-          if [[ -f "package.json" ]]; then
-            UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
-            # Exclude dependencies used in scripts, code, and workspace packages
-            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat root_used_deps.txt root_used_code.txt root_workspace_deps.txt | sort) || echo "")
-            echo "ROOT_UNUSED<<EOF" >> $GITHUB_ENV
-            echo "$UNUSED" >> $GITHUB_ENV
-            echo "EOF" >> $GITHUB_ENV
-          fi
-
-      - name: Run depcheck for client/package.json
-        id: check-client
-        run: |
-          if [[ -f "client/package.json" ]]; then
-            chmod -R 755 client
-            cd client
-            UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
-            # Exclude dependencies used in scripts, code, and workspace packages
-            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../client_used_deps.txt ../client_used_code.txt ../client_workspace_deps.txt | sort) || echo "")
-            # Filter out false positives
-            UNUSED=$(echo "$UNUSED" | grep -v "^micromark-extension-llm-math$" || echo "")
-            echo "CLIENT_UNUSED<<EOF" >> $GITHUB_ENV
-            echo "$UNUSED" >> $GITHUB_ENV
-            echo "EOF" >> $GITHUB_ENV
-            cd ..
-          fi
-
-      - name: Run depcheck for api/package.json
-        id: check-api
-        run: |
-          if [[ -f "api/package.json" ]]; then
-            chmod -R 755 api
-            cd api
-            UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
-            # Exclude dependencies used in scripts, code, and workspace packages
-            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../api_used_deps.txt ../api_used_code.txt ../api_workspace_deps.txt | sort) || echo "")
-            echo "API_UNUSED<<EOF" >> $GITHUB_ENV
-            echo "$UNUSED" >> $GITHUB_ENV
-            echo "EOF" >> $GITHUB_ENV
-            cd ..
-          fi
-
-      - name: Post comment on PR if unused dependencies are found
-        if: env.ROOT_UNUSED != '' || env.CLIENT_UNUSED != '' || env.API_UNUSED != ''
-        run: |
-          PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH")
-
-          ROOT_LIST=$(echo "$ROOT_UNUSED" | awk '{print "- `" $0 "`"}')
-          CLIENT_LIST=$(echo "$CLIENT_UNUSED" | awk '{print "- `" $0 "`"}')
-          API_LIST=$(echo "$API_UNUSED" | awk '{print "- `" $0 "`"}')
-
-          COMMENT_BODY=$(cat <<EOF
-          ### 🚨 Unused NPM Packages Detected
-
-          The following **unused dependencies** were found:
-
-          $(if [[ ! -z "$ROOT_UNUSED" ]]; then echo "#### 📂 Root \`package.json\`"; echo ""; echo "$ROOT_LIST"; echo ""; fi)
-
-          $(if [[ ! -z "$CLIENT_UNUSED" ]]; then echo "#### 📂 Client \`client/package.json\`"; echo ""; echo "$CLIENT_LIST"; echo ""; fi)
-
-          $(if [[ ! -z "$API_UNUSED" ]]; then echo "#### 📂 API \`api/package.json\`"; echo ""; echo "$API_LIST"; echo ""; fi)
-
-          ⚠️ **Please remove these unused dependencies to keep your project clean.**
-          EOF
-          )
-
-          gh api "repos/${{ github.repository }}/issues/${PR_NUMBER}/comments" \
-            -f body="$COMMENT_BODY" \
-            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}"
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Fail workflow if unused dependencies found
-        if: env.ROOT_UNUSED != '' || env.CLIENT_UNUSED != '' || env.API_UNUSED != ''
-        run: exit 1
--- a/.gitignore
+++ b/.gitignore
@@ -13,9 +13,6 @@ pids
 *.seed
 .git

-# CI/CD data
-test-image*
-
 # Directory for instrumented libs generated by jscoverage/JSCover
 lib-cov

@@ -40,10 +37,6 @@ client/public/main.js
 client/public/main.js.map
 client/public/main.js.LICENSE.txt

-# Azure Blob Storage Emulator (Azurite)
-__azurite**
-__blobstorage__/**/*
-
 # Dependency directorys
 # Deployed apps should consider commenting these lines out:
 # see https://npmjs.org/doc/faq.html#Should-I-check-my-node_modules-folder-into-git
@@ -55,11 +48,6 @@ bower_components/
 *.d.ts
 !vite-env.d.ts

-# AI
-.clineignore
-.cursor
-.aider*
-
 # Floobits
 .floo
 .floobit
@@ -117,24 +105,4 @@ auth.json
 uploads/

 # owner
-release/
-
-# Helm
-helm/librechat/Chart.lock
-helm/**/charts/
-helm/**/.values.yaml
-
-!/client/src/@types/i18next.d.ts
-
-# SAML Idp cert
-*.cert
-
-# AI Assistants
-/.claude/
-/.cursor/
-/.copilot/
-/.aider/
-/.openai/
-/.tabnine/
-/.codeium
-*.local.md
+release/
--- a/.prettierrc
+++ b/.prettierrc
@@ -1,19 +0,0 @@
-{
-  "tailwindConfig": "./client/tailwind.config.cjs",
-  "printWidth": 100,
-  "tabWidth": 2,
-  "useTabs": false,
-  "semi": true,
-  "singleQuote": true,
-  "trailingComma": "all",
-  "arrowParens": "always",
-  "embeddedLanguageFormatting": "auto",
-  "insertPragma": false,
-  "proseWrap": "preserve",
-  "quoteProps": "as-needed",
-  "requirePragma": false,
-  "rangeStart": 0,
-  "endOfLine": "auto",
-  "jsxSingleQuote": false,
-  "plugins": ["prettier-plugin-tailwindcss"]
-}
--- a/.vscode/launch.json
+++ b/.vscode/launch.json
@@ -8,11 +8,9 @@
      "skipFiles": ["<node_internals>/**"],
      "program": "${workspaceFolder}/api/server/index.js",
      "env": {
-        "NODE_ENV": "production",
-        "NODE_TLS_REJECT_UNAUTHORIZED": "0"
+        "NODE_ENV": "production"
      },
-      "console": "integratedTerminal",
-      "envFile": "${workspaceFolder}/.env"
+      "console": "integratedTerminal"
    }
  ]
 }
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,236 +0,0 @@
-# Changelog
-
-All notable changes to this project will be documented in this file.
-
-
-
-
-
-
-## [Unreleased]
-
-### ✨ New Features
-
- ✨ feat: implement search parameter updates by **@mawburn** in [#7151](https://github.com/danny-avila/LibreChat/pull/7151)
- 🎏 feat: Add MCP support for Streamable HTTP Transport by **@benverhees** in [#7353](https://github.com/danny-avila/LibreChat/pull/7353)
- 🔒 feat: Add Content Security Policy using Helmet middleware by **@rubentalstra** in [#7377](https://github.com/danny-avila/LibreChat/pull/7377)
- ✨ feat: Add Normalization for MCP Server Names by **@danny-avila** in [#7421](https://github.com/danny-avila/LibreChat/pull/7421)
- 📊 feat: Improve Helm Chart by **@hofq** in [#3638](https://github.com/danny-avila/LibreChat/pull/3638)
- 🦾 feat: Claude-4 Support by **@danny-avila** in [#7509](https://github.com/danny-avila/LibreChat/pull/7509)
- 🪨 feat: Bedrock Support for Claude-4 Reasoning by **@danny-avila** in [#7517](https://github.com/danny-avila/LibreChat/pull/7517)
-
-### 🌍 Internationalization
-
- 🌍 i18n: Add `Danish` and `Czech` and `Catalan` localization support by **@rubentalstra** in [#7373](https://github.com/danny-avila/LibreChat/pull/7373)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7375](https://github.com/danny-avila/LibreChat/pull/7375)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7468](https://github.com/danny-avila/LibreChat/pull/7468)
-
-### 🔧 Fixes
-
- 💬 fix: update aria-label for accessibility in ConvoLink component by **@berry-13** in [#7320](https://github.com/danny-avila/LibreChat/pull/7320)
- 🔑 fix: use `apiKey` instead of `openAIApiKey` in OpenAI-like Config by **@danny-avila** in [#7337](https://github.com/danny-avila/LibreChat/pull/7337)
- 🔄 fix: update navigation logic in `useFocusChatEffect` to ensure correct search parameters are used by **@mawburn** in [#7340](https://github.com/danny-avila/LibreChat/pull/7340)
- 🔄 fix: Improve MCP Connection Cleanup by **@danny-avila** in [#7400](https://github.com/danny-avila/LibreChat/pull/7400)
- 🛡️ fix: Preset and Validation Logic for URL Query Params by **@danny-avila** in [#7407](https://github.com/danny-avila/LibreChat/pull/7407)
- 🌘 fix: artifact of preview text is illegible in dark mode by **@nhtruong** in [#7405](https://github.com/danny-avila/LibreChat/pull/7405)
- 🛡️ fix: Temporarily Remove CSP until Configurable by **@danny-avila** in [#7419](https://github.com/danny-avila/LibreChat/pull/7419)
- 💽 fix: Exclude index page `/` from static cache settings by **@sbruel** in [#7382](https://github.com/danny-avila/LibreChat/pull/7382)
-
-### ⚙️ Other Changes
-
- 📜 docs: CHANGELOG for release v0.7.8 by **@github-actions[bot]** in [#7290](https://github.com/danny-avila/LibreChat/pull/7290)
- 📦 chore: Update API Package Dependencies by **@danny-avila** in [#7359](https://github.com/danny-avila/LibreChat/pull/7359)
- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7321](https://github.com/danny-avila/LibreChat/pull/7321)
- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7434](https://github.com/danny-avila/LibreChat/pull/7434)
- 🛡️ chore: `multer` v2.0.0 for CVE-2025-47935 and CVE-2025-47944 by **@danny-avila** in [#7454](https://github.com/danny-avila/LibreChat/pull/7454)
- 📂 refactor: Improve `FileAttachment` & File Form Deletion by **@danny-avila** in [#7471](https://github.com/danny-avila/LibreChat/pull/7471)
- 📊 chore: Remove Old Helm Chart by **@hofq** in [#7512](https://github.com/danny-avila/LibreChat/pull/7512)
- 🪖 chore: bump helm app version to v0.7.8 by **@austin-barrington** in [#7524](https://github.com/danny-avila/LibreChat/pull/7524)
-
-
-
---
-## [v0.7.8] - 
-
-Changes from v0.7.8-rc1 to v0.7.8.
-
-### ✨ New Features
-
- ✨ feat: Enhance form submission for touch screens by **@berry-13** in [#7198](https://github.com/danny-avila/LibreChat/pull/7198)
- 🔍 feat: Additional Tavily API Tool Parameters by **@glowforge-opensource** in [#7232](https://github.com/danny-avila/LibreChat/pull/7232)
- 🐋 feat: Add python to Dockerfile for increased MCP compatibility by **@technicalpickles** in [#7270](https://github.com/danny-avila/LibreChat/pull/7270)
-
-### 🔧 Fixes
-
- 🔧 fix: Google Gemma Support & OpenAI Reasoning Instructions by **@danny-avila** in [#7196](https://github.com/danny-avila/LibreChat/pull/7196)
- 🛠️ fix: Conversation Navigation State by **@danny-avila** in [#7210](https://github.com/danny-avila/LibreChat/pull/7210)
- 🔄 fix: o-Series Model Regex for System Messages by **@danny-avila** in [#7245](https://github.com/danny-avila/LibreChat/pull/7245)
- 🔖 fix: Custom Headers for Initial MCP SSE Connection by **@danny-avila** in [#7246](https://github.com/danny-avila/LibreChat/pull/7246)
- 🛡️ fix: Deep Clone `MCPOptions` for User MCP Connections by **@danny-avila** in [#7247](https://github.com/danny-avila/LibreChat/pull/7247)
- 🔄 fix: URL Param Race Condition and File Draft Persistence by **@danny-avila** in [#7257](https://github.com/danny-avila/LibreChat/pull/7257)
- 🔄 fix: Assistants Endpoint & Minor Issues by **@danny-avila** in [#7274](https://github.com/danny-avila/LibreChat/pull/7274)
- 🔄 fix: Ollama Think Tag Edge Case with Tools by **@danny-avila** in [#7275](https://github.com/danny-avila/LibreChat/pull/7275)
-
-### ⚙️ Other Changes
-
- 📜 docs: CHANGELOG for release v0.7.8-rc1 by **@github-actions[bot]** in [#7153](https://github.com/danny-avila/LibreChat/pull/7153)
- 🔄 refactor: Artifact Visibility Management by **@danny-avila** in [#7181](https://github.com/danny-avila/LibreChat/pull/7181)
- 📦 chore: Bump Package Security by **@danny-avila** in [#7183](https://github.com/danny-avila/LibreChat/pull/7183)
- 🌿 refactor: Unmount Fork Popover on Hide for Better Performance by **@danny-avila** in [#7189](https://github.com/danny-avila/LibreChat/pull/7189)
- 🧰 chore: ESLint configuration to enforce Prettier formatting rules by **@mawburn** in [#7186](https://github.com/danny-avila/LibreChat/pull/7186)
- 🎨 style: Improve KaTeX Rendering for LaTeX Equations by **@andresgit** in [#7223](https://github.com/danny-avila/LibreChat/pull/7223)
- 📝 docs: Update `.env.example` Google models by **@marlonka** in [#7254](https://github.com/danny-avila/LibreChat/pull/7254)
- 💬 refactor: MCP Chat Visibility Option, Google Rates, Remove OpenAPI Plugins by **@danny-avila** in [#7286](https://github.com/danny-avila/LibreChat/pull/7286)
- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7214](https://github.com/danny-avila/LibreChat/pull/7214)
-
-
-
-[See full release details][release-v0.7.8]
-
-[release-v0.7.8]: https://github.com/danny-avila/LibreChat/releases/tag/v0.7.8
-
---
-## [v0.7.8-rc1] - 
-
-Changes from v0.7.7 to v0.7.8-rc1.
-
-### ✨ New Features
-
- 🔍 feat: Mistral OCR API / Upload Files as Text by **@danny-avila** in [#6274](https://github.com/danny-avila/LibreChat/pull/6274)
- 🤖 feat: Support OpenAI Web Search models by **@danny-avila** in [#6313](https://github.com/danny-avila/LibreChat/pull/6313)
- 🔗 feat: Agent Chain (Mixture-of-Agents) by **@danny-avila** in [#6374](https://github.com/danny-avila/LibreChat/pull/6374)
- ⌛ feat: `initTimeout` for Slow Starting MCP Servers by **@perweij** in [#6383](https://github.com/danny-avila/LibreChat/pull/6383)
- 🚀 feat: `S3` Integration for File handling and Image uploads by **@rubentalstra** in [#6142](https://github.com/danny-avila/LibreChat/pull/6142)
- 🔒feat: Enable OpenID Auto-Redirect by **@leondape** in [#6066](https://github.com/danny-avila/LibreChat/pull/6066)
- 🚀 feat: Integrate `Azure Blob Storage` for file handling and image uploads by **@rubentalstra** in [#6153](https://github.com/danny-avila/LibreChat/pull/6153)
- 🚀 feat: Add support for custom `AWS` endpoint in `S3` by **@rubentalstra** in [#6431](https://github.com/danny-avila/LibreChat/pull/6431)
- 🚀 feat: Add support for LDAP STARTTLS in LDAP authentication by **@rubentalstra** in [#6438](https://github.com/danny-avila/LibreChat/pull/6438)
- 🚀 feat: Refactor schema exports and update package version to 0.0.4 by **@rubentalstra** in [#6455](https://github.com/danny-avila/LibreChat/pull/6455)
- 🔼 feat: Add Auto Submit For URL Query Params by **@mjaverto** in [#6440](https://github.com/danny-avila/LibreChat/pull/6440)
- 🛠 feat: Enhance Redis Integration, Rate Limiters & Log Headers by **@danny-avila** in [#6462](https://github.com/danny-avila/LibreChat/pull/6462)
- 💵 feat: Add Automatic Balance Refill by **@rubentalstra** in [#6452](https://github.com/danny-avila/LibreChat/pull/6452)
- 🗣️ feat: add support for gpt-4o-transcribe models by **@berry-13** in [#6483](https://github.com/danny-avila/LibreChat/pull/6483)
- 🎨 feat: UI Refresh for Enhanced UX by **@berry-13** in [#6346](https://github.com/danny-avila/LibreChat/pull/6346)
- 🌍 feat: Add support for Hungarian language localization by **@rubentalstra** in [#6508](https://github.com/danny-avila/LibreChat/pull/6508)
- 🚀 feat: Add Gemini 2.5 Token/Context Values, Increase Max Possible Output to 64k by **@danny-avila** in [#6563](https://github.com/danny-avila/LibreChat/pull/6563)
- 🚀 feat: Enhance MCP Connections For Multi-User Support by **@danny-avila** in [#6610](https://github.com/danny-avila/LibreChat/pull/6610)
- 🚀 feat: Enhance S3 URL Expiry with Refresh; fix: S3 File Deletion by **@danny-avila** in [#6647](https://github.com/danny-avila/LibreChat/pull/6647)
- 🚀 feat: enhance UI components and refactor settings by **@berry-13** in [#6625](https://github.com/danny-avila/LibreChat/pull/6625)
- 💬 feat: move TemporaryChat to the Header by **@berry-13** in [#6646](https://github.com/danny-avila/LibreChat/pull/6646)
- 🚀 feat: Use Model Specs + Specific Endpoints, Limit Providers for Agents by **@danny-avila** in [#6650](https://github.com/danny-avila/LibreChat/pull/6650)
- 🪙 feat: Sync Balance Config on Login by **@danny-avila** in [#6671](https://github.com/danny-avila/LibreChat/pull/6671)
- 🔦 feat: MCP Support for Non-Agent Endpoints by **@danny-avila** in [#6775](https://github.com/danny-avila/LibreChat/pull/6775)
- 🗃️ feat: Code Interpreter File Persistence between Sessions by **@danny-avila** in [#6790](https://github.com/danny-avila/LibreChat/pull/6790)
- 🖥️ feat: Code Interpreter API for Non-Agent Endpoints by **@danny-avila** in [#6803](https://github.com/danny-avila/LibreChat/pull/6803)
- ⚡ feat: Self-hosted Artifacts Static Bundler URL by **@danny-avila** in [#6827](https://github.com/danny-avila/LibreChat/pull/6827)
- 🐳 feat: Add Jemalloc and UV to Docker Builds by **@danny-avila** in [#6836](https://github.com/danny-avila/LibreChat/pull/6836)
- 🤖 feat: GPT-4.1 by **@danny-avila** in [#6880](https://github.com/danny-avila/LibreChat/pull/6880)
- 👋 feat: remove Edge TTS by **@berry-13** in [#6885](https://github.com/danny-avila/LibreChat/pull/6885)
- feat: nav optimization  by **@berry-13** in [#5785](https://github.com/danny-avila/LibreChat/pull/5785)
- 🗺️ feat: Add Parameter Location Mapping for OpenAPI actions by **@peeeteeer** in [#6858](https://github.com/danny-avila/LibreChat/pull/6858)
- 🤖 feat: Support `o4-mini` and `o3` Models by **@danny-avila** in [#6928](https://github.com/danny-avila/LibreChat/pull/6928)
- 🎨 feat: OpenAI Image Tools (GPT-Image-1) by **@danny-avila** in [#7079](https://github.com/danny-avila/LibreChat/pull/7079)
- 🗓️ feat: Add Special Variables for Prompts & Agents, Prompt UI Improvements by **@danny-avila** in [#7123](https://github.com/danny-avila/LibreChat/pull/7123)
-
-### 🌍 Internationalization
-
- 🌍 i18n: Add Thai Language Support and Update Translations by **@rubentalstra** in [#6219](https://github.com/danny-avila/LibreChat/pull/6219)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6220](https://github.com/danny-avila/LibreChat/pull/6220)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6240](https://github.com/danny-avila/LibreChat/pull/6240)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6241](https://github.com/danny-avila/LibreChat/pull/6241)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6277](https://github.com/danny-avila/LibreChat/pull/6277)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6414](https://github.com/danny-avila/LibreChat/pull/6414)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6505](https://github.com/danny-avila/LibreChat/pull/6505)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6530](https://github.com/danny-avila/LibreChat/pull/6530)
- 🌍 i18n: Add Persian Localization Support by **@rubentalstra** in [#6669](https://github.com/danny-avila/LibreChat/pull/6669)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#6667](https://github.com/danny-avila/LibreChat/pull/6667)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7126](https://github.com/danny-avila/LibreChat/pull/7126)
- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7148](https://github.com/danny-avila/LibreChat/pull/7148)
-
-### 👐 Accessibility
-
- 🎨 a11y: Update Model Spec Description Text by **@berry-13** in [#6294](https://github.com/danny-avila/LibreChat/pull/6294)
- 🗑️ a11y: Add Accessible Name to Button for File Attachment Removal by **@kangabell** in [#6709](https://github.com/danny-avila/LibreChat/pull/6709)
- ⌨️ a11y: enhance accessibility & visual consistency by **@berry-13** in [#6866](https://github.com/danny-avila/LibreChat/pull/6866)
- 🙌 a11y: Searchbar/Conversations List Focus by **@danny-avila** in [#7096](https://github.com/danny-avila/LibreChat/pull/7096)
- 👐 a11y: Improve Fork and SplitText Accessibility by **@danny-avila** in [#7147](https://github.com/danny-avila/LibreChat/pull/7147)
-
-### 🔧 Fixes
-
- 🐛 fix: Avatar Type Definitions in Agent/Assistant Schemas by **@danny-avila** in [#6235](https://github.com/danny-avila/LibreChat/pull/6235)
- 🔧 fix: MeiliSearch Field Error and Patch Incorrect Import by #6210 by **@rubentalstra** in [#6245](https://github.com/danny-avila/LibreChat/pull/6245)
- 🔏 fix: Enhance Two-Factor Authentication by **@rubentalstra** in [#6247](https://github.com/danny-avila/LibreChat/pull/6247)
- 🐛 fix: Await saveMessage in abortMiddleware to ensure proper execution by **@sh4shii** in [#6248](https://github.com/danny-avila/LibreChat/pull/6248)
- 🔧 fix: Axios Proxy Usage And Bump `mongoose` by **@danny-avila** in [#6298](https://github.com/danny-avila/LibreChat/pull/6298)
- 🔧 fix: comment out MCP servers to resolve service run issues by **@KunalScriptz** in [#6316](https://github.com/danny-avila/LibreChat/pull/6316)
- 🔧 fix: Update Token Calculations and Mapping, MCP `env` Initialization by **@danny-avila** in [#6406](https://github.com/danny-avila/LibreChat/pull/6406)
- 🐞 fix: Agent "Resend" Message Attachments + Source Icon Styling by **@danny-avila** in [#6408](https://github.com/danny-avila/LibreChat/pull/6408)
- 🐛 fix: Prevent Crash on Duplicate Message ID by **@Odrec** in [#6392](https://github.com/danny-avila/LibreChat/pull/6392)
- 🔐 fix: Invalid Key Length in 2FA Encryption by **@rubentalstra** in [#6432](https://github.com/danny-avila/LibreChat/pull/6432)
- 🏗️ fix: Fix Agents Token Spend Race Conditions, Expand Test Coverage by **@danny-avila** in [#6480](https://github.com/danny-avila/LibreChat/pull/6480)
- 🔃 fix: Draft Clearing, Claude Titles, Remove Default Vision Max Tokens by **@danny-avila** in [#6501](https://github.com/danny-avila/LibreChat/pull/6501)
- 🔧 fix: Update username reference to use user.name in greeting display by **@rubentalstra** in [#6534](https://github.com/danny-avila/LibreChat/pull/6534)
- 🔧 fix: S3 Download Stream with Key Extraction and Blob Storage Encoding for Vision by **@danny-avila** in [#6557](https://github.com/danny-avila/LibreChat/pull/6557)
- 🔧 fix: Mistral type strictness for `usage` & update token values/windows by **@danny-avila** in [#6562](https://github.com/danny-avila/LibreChat/pull/6562)
- 🔧 fix: Consolidate Text Parsing and TTS Edge Initialization by **@danny-avila** in [#6582](https://github.com/danny-avila/LibreChat/pull/6582)
- 🔧 fix: Ensure continuation in image processing on base64 encoding from Blob Storage by **@danny-avila** in [#6619](https://github.com/danny-avila/LibreChat/pull/6619)
- ✉️ fix: Fallback For User Name In Email Templates by **@danny-avila** in [#6620](https://github.com/danny-avila/LibreChat/pull/6620)
- 🔧 fix: Azure Blob Integration and File Source References by **@rubentalstra** in [#6575](https://github.com/danny-avila/LibreChat/pull/6575)
- 🐛 fix: Safeguard against undefined addedEndpoints by **@wipash** in [#6654](https://github.com/danny-avila/LibreChat/pull/6654)
- 🤖 fix: Gemini 2.5 Vision Support by **@danny-avila** in [#6663](https://github.com/danny-avila/LibreChat/pull/6663)
- 🔄 fix: Avatar & Error Handling Enhancements by **@danny-avila** in [#6687](https://github.com/danny-avila/LibreChat/pull/6687)
- 🔧 fix: Chat Middleware, Zod Conversion, Auto-Save and S3 URL Refresh by **@danny-avila** in [#6720](https://github.com/danny-avila/LibreChat/pull/6720)
- 🔧 fix: Agent Capability Checks & DocumentDB Compatibility for Agent Resource Removal by **@danny-avila** in [#6726](https://github.com/danny-avila/LibreChat/pull/6726)
- 🔄 fix: Improve audio MIME type detection and handling by **@berry-13** in [#6707](https://github.com/danny-avila/LibreChat/pull/6707)
- 🪺 fix: Update Role Handling due to New Schema Shape by **@danny-avila** in [#6774](https://github.com/danny-avila/LibreChat/pull/6774)
- 🗨️ fix: Show ModelSpec Greeting by **@berry-13** in [#6770](https://github.com/danny-avila/LibreChat/pull/6770)
- 🔧 fix: Keyv and Proxy Issues, and More Memory Optimizations by **@danny-avila** in [#6867](https://github.com/danny-avila/LibreChat/pull/6867)
- ✨ fix: Implement dynamic text sizing for greeting and name display by **@berry-13** in [#6833](https://github.com/danny-avila/LibreChat/pull/6833)
- 📝 fix: Mistral OCR Image Support and Azure Agent Titles by **@danny-avila** in [#6901](https://github.com/danny-avila/LibreChat/pull/6901)
- 📢 fix: Invalid `engineTTS` and Conversation State on Navigation by **@berry-13** in [#6904](https://github.com/danny-avila/LibreChat/pull/6904)
- 🛠️ fix: Improve Accessibility and Display of Conversation Menu by **@danny-avila** in [#6913](https://github.com/danny-avila/LibreChat/pull/6913)
- 🔧 fix: Agent Resource Form, Convo Menu Style, Ensure Draft Clears on Submission by **@danny-avila** in [#6925](https://github.com/danny-avila/LibreChat/pull/6925)
- 🔀 fix: MCP Improvements, Auto-Save Drafts, Artifact Markup by **@danny-avila** in [#7040](https://github.com/danny-avila/LibreChat/pull/7040)
- 🐋 fix: Improve Deepseek Compatbility by **@danny-avila** in [#7132](https://github.com/danny-avila/LibreChat/pull/7132)
- 🐙 fix: Add Redis Ping Interval to Prevent Connection Drops by **@peeeteeer** in [#7127](https://github.com/danny-avila/LibreChat/pull/7127)
-
-### ⚙️ Other Changes
-
- 📦 refactor: Move DB Models to `@librechat/data-schemas` by **@rubentalstra** in [#6210](https://github.com/danny-avila/LibreChat/pull/6210)
- 📦 chore: Patch `axios` to address CVE-2025-27152 by **@danny-avila** in [#6222](https://github.com/danny-avila/LibreChat/pull/6222)
- ⚠️ refactor: Use Error Content Part Instead Of Throwing Error for Agents by **@danny-avila** in [#6262](https://github.com/danny-avila/LibreChat/pull/6262)
- 🏃‍♂️ refactor: Improve Agent Run Context & Misc. Changes by **@danny-avila** in [#6448](https://github.com/danny-avila/LibreChat/pull/6448)
- 📝 docs: librechat.example.yaml by **@ineiti** in [#6442](https://github.com/danny-avila/LibreChat/pull/6442)
- 🏃‍♂️ refactor: More Agent Context Improvements during Run by **@danny-avila** in [#6477](https://github.com/danny-avila/LibreChat/pull/6477)
- 🔃 refactor: Allow streaming for `o1` models by **@danny-avila** in [#6509](https://github.com/danny-avila/LibreChat/pull/6509)
- 🔧 chore: `Vite` Plugin Upgrades & Config Optimizations by **@rubentalstra** in [#6547](https://github.com/danny-avila/LibreChat/pull/6547)
- 🔧 refactor: Consolidate Logging, Model Selection & Actions Optimizations, Minor Fixes by **@danny-avila** in [#6553](https://github.com/danny-avila/LibreChat/pull/6553)
- 🎨 style: Address Minor UI Refresh Issues by **@berry-13** in [#6552](https://github.com/danny-avila/LibreChat/pull/6552)
- 🔧 refactor: Enhance Model & Endpoint Configurations with Global Indicators 🌍 by **@berry-13** in [#6578](https://github.com/danny-avila/LibreChat/pull/6578)
- 💬 style: Chat UI, Greeting, and Message adjustments by **@berry-13** in [#6612](https://github.com/danny-avila/LibreChat/pull/6612)
- ⚡ refactor: DocumentDB Compatibility for Balance Updates by **@danny-avila** in [#6673](https://github.com/danny-avila/LibreChat/pull/6673)
- 🧹 chore: Update ESLint rules for React hooks by **@rubentalstra** in [#6685](https://github.com/danny-avila/LibreChat/pull/6685)
- 🪙 chore: Update Gemini Pricing by **@RedwindA** in [#6731](https://github.com/danny-avila/LibreChat/pull/6731)
- 🪺 refactor: Nest Permission fields for Roles by **@rubentalstra** in [#6487](https://github.com/danny-avila/LibreChat/pull/6487)
- 📦 chore: Update `caniuse-lite` dependency to version 1.0.30001706 by **@rubentalstra** in [#6482](https://github.com/danny-avila/LibreChat/pull/6482)
- ⚙️ refactor: OAuth Flow Signal, Type Safety, Tool Progress & Updated Packages by **@danny-avila** in [#6752](https://github.com/danny-avila/LibreChat/pull/6752)
- 📦 chore: bump vite from 6.2.3 to 6.2.5 by **@dependabot[bot]** in [#6745](https://github.com/danny-avila/LibreChat/pull/6745)
- 💾 chore: Enhance Local Storage Handling and Update MCP SDK by **@danny-avila** in [#6809](https://github.com/danny-avila/LibreChat/pull/6809)
- 🤖 refactor: Improve Agents Memory Usage, Bump Keyv, Grok 3 by **@danny-avila** in [#6850](https://github.com/danny-avila/LibreChat/pull/6850)
- 💾 refactor: Enhance Memory In Image Encodings & Client Disposal by **@danny-avila** in [#6852](https://github.com/danny-avila/LibreChat/pull/6852)
- 🔁 refactor: Token Event Handler and Standardize `maxTokens` Key by **@danny-avila** in [#6886](https://github.com/danny-avila/LibreChat/pull/6886)
- 🔍 refactor: Search & Message Retrieval by **@berry-13** in [#6903](https://github.com/danny-avila/LibreChat/pull/6903)
- 🎨 style: standardize dropdown styling & fix z-Index layering by **@berry-13** in [#6939](https://github.com/danny-avila/LibreChat/pull/6939)
- 📙 docs: CONTRIBUTING.md by **@dblock** in [#6831](https://github.com/danny-avila/LibreChat/pull/6831)
- 🧭 refactor: Modernize Nav/Header by **@danny-avila** in [#7094](https://github.com/danny-avila/LibreChat/pull/7094)
- 🪶 refactor: Chat Input Focus for Conversation Navigations & ChatForm Optimizations by **@danny-avila** in [#7100](https://github.com/danny-avila/LibreChat/pull/7100)
- 🔃 refactor: Streamline Navigation, Message Loading UX by **@danny-avila** in [#7118](https://github.com/danny-avila/LibreChat/pull/7118)
- 📜 docs: Unreleased changelog by **@github-actions[bot]** in [#6265](https://github.com/danny-avila/LibreChat/pull/6265)
-
-
-
-[See full release details][release-v0.7.8-rc1]
-
-[release-v0.7.8-rc1]: https://github.com/danny-avila/LibreChat/releases/tag/v0.7.8-rc1
-
---
--- a/30
+++ b/30
@@ -1,49 +1,33 @@
-# v0.8.0-rc4
+# v0.7.5

 # Base node image
 FROM node:20-alpine AS node

-# Install jemalloc
-RUN apk add --no-cache jemalloc
-RUN apk add --no-cache python3 py3-pip uv
-
-# Set environment variable to use jemalloc
-ENV LD_PRELOAD=/usr/lib/libjemalloc.so.2
-
-# Add `uv` for extended MCP support
-COPY --from=ghcr.io/astral-sh/uv:0.6.13 /uv /uvx /bin/
-RUN uv --version
+RUN apk --no-cache add curl

 RUN mkdir -p /app && chown node:node /app
 WORKDIR /app

 USER node

-COPY --chown=node:node package.json package-lock.json ./
-COPY --chown=node:node api/package.json ./api/package.json
-COPY --chown=node:node client/package.json ./client/package.json
-COPY --chown=node:node packages/data-provider/package.json ./packages/data-provider/package.json
-COPY --chown=node:node packages/data-schemas/package.json ./packages/data-schemas/package.json
-COPY --chown=node:node packages/api/package.json ./packages/api/package.json
+COPY --chown=node:node . .

 RUN \
    # Allow mounting of these files, which have no default
    touch .env ; \
    # Create directories for the volumes to inherit the correct permissions
-    mkdir -p /app/client/public/images /app/api/logs /app/uploads ; \
+    mkdir -p /app/client/public/images /app/api/logs ; \
    npm config set fetch-retry-maxtimeout 600000 ; \
    npm config set fetch-retries 5 ; \
    npm config set fetch-retry-mintimeout 15000 ; \
-    npm ci --no-audit
-
-COPY --chown=node:node . .
-
-RUN \
+    npm install --no-audit; \
    # React client build
    NODE_OPTIONS="--max-old-space-size=2048" npm run frontend; \
    npm prune --production; \
    npm cache clean --force

+RUN mkdir -p /app/client/public/images /app/api/logs
+
 # Node API setup
 EXPOSE 3080
 ENV HOST=0.0.0.0
--- a/Dockerfile.multi
+++ b/Dockerfile.multi
@@ -1,12 +1,8 @@
 # Dockerfile.multi
-# v0.8.0-rc4
+# v0.7.5

 # Base for all builds
-FROM node:20-alpine AS base-min
-# Install jemalloc
-RUN apk add --no-cache jemalloc
-# Set environment variable to use jemalloc
-ENV LD_PRELOAD=/usr/lib/libjemalloc.so.2
+FROM node:20-alpine AS base
 WORKDIR /app
 RUN apk --no-cache add curl
 RUN npm config set fetch-retry-maxtimeout 600000 && \
@@ -14,69 +10,35 @@ RUN npm config set fetch-retry-maxtimeout 600000 && \
    npm config set fetch-retry-mintimeout 15000
 COPY package*.json ./
 COPY packages/data-provider/package*.json ./packages/data-provider/
-COPY packages/api/package*.json ./packages/api/
-COPY packages/data-schemas/package*.json ./packages/data-schemas/
-COPY packages/client/package*.json ./packages/client/
 COPY client/package*.json ./client/
 COPY api/package*.json ./api/
-
-# Install all dependencies for every build
-FROM base-min AS base
-WORKDIR /app
 RUN npm ci

-# Build `data-provider` package
+# Build data-provider
 FROM base AS data-provider-build
 WORKDIR /app/packages/data-provider
 COPY packages/data-provider ./
 RUN npm run build
-
-# Build `data-schemas` package
-FROM base AS data-schemas-build
-WORKDIR /app/packages/data-schemas
-COPY packages/data-schemas ./
-COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
-RUN npm run build
-
-# Build `api` package
-FROM base AS api-package-build
-WORKDIR /app/packages/api
-COPY packages/api ./
-COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
-COPY --from=data-schemas-build /app/packages/data-schemas/dist /app/packages/data-schemas/dist
-RUN npm run build
-
-# Build `client` package
-FROM base AS client-package-build
-WORKDIR /app/packages/client
-COPY packages/client ./
-RUN npm run build
+RUN npm prune --production

 # Client build
 FROM base AS client-build
 WORKDIR /app/client
 COPY client ./
 COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
-COPY --from=client-package-build /app/packages/client/dist /app/packages/client/dist
-COPY --from=client-package-build /app/packages/client/src /app/packages/client/src
 ENV NODE_OPTIONS="--max-old-space-size=2048"
 RUN npm run build
+RUN npm prune --production

 # API setup (including client dist)
-FROM base-min AS api-build
-# Add `uv` for extended MCP support
-COPY --from=ghcr.io/astral-sh/uv:0.6.13 /uv /uvx /bin/
-RUN uv --version
+FROM base AS api-build
 WORKDIR /app
-# Install only production deps
-RUN npm ci --omit=dev
 COPY api ./api
 COPY config ./config
 COPY --from=data-provider-build /app/packages/data-provider/dist ./packages/data-provider/dist
-COPY --from=data-schemas-build /app/packages/data-schemas/dist ./packages/data-schemas/dist
-COPY --from=api-package-build /app/packages/api/dist ./packages/api/dist
 COPY --from=client-build /app/client/dist ./client/dist
 WORKDIR /app/api
+RUN npm prune --production
 EXPOSE 3080
 ENV HOST=0.0.0.0
-CMD ["node", "server/index.js"]
+CMD ["node", "server/index.js"]
--- a/2
+++ b/2
@@ -1,6 +1,6 @@
 MIT License

-Copyright (c) 2025 LibreChat
+Copyright (c) 2024 LibreChat

 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal
--- a/README.md
+++ b/README.md
@@ -38,100 +38,42 @@
 </a>
 </p>

-<p align="center">
-  <a href="https://www.librechat.ai/docs/translation">
-    <img 
-      src="https://img.shields.io/badge/dynamic/json.svg?style=for-the-badge&color=2096F3&label=locize&query=%24.translatedPercentage&url=https://api.locize.app/badgedata/4cb2598b-ed4d-469c-9b04-2ed531a8cb45&suffix=%+translated" 
-      alt="Translation Progress">
-  </a>
-</p>
+# 📃 Features

-
-# ✨ Features
-
- 🖥️ **UI & Experience** inspired by ChatGPT with enhanced design and features
-
- 🤖 **AI Model Selection**:  
-  - Anthropic (Claude), AWS Bedrock, OpenAI, Azure OpenAI, Google, Vertex AI, OpenAI Responses API (incl. Azure)
-  - [Custom Endpoints](https://www.librechat.ai/docs/quick_start/custom_endpoints): Use any OpenAI-compatible API with LibreChat, no proxy required
-  - Compatible with [Local & Remote AI Providers](https://www.librechat.ai/docs/configuration/librechat_yaml/ai_endpoints):
-    - Ollama, groq, Cohere, Mistral AI, Apple MLX, koboldcpp, together.ai,
-    - OpenRouter, Perplexity, ShuttleAI, Deepseek, Qwen, and more
-
- 🔧 **[Code Interpreter API](https://www.librechat.ai/docs/features/code_interpreter)**: 
-  - Secure, Sandboxed Execution in Python, Node.js (JS/TS), Go, C/C++, Java, PHP, Rust, and Fortran
-  - Seamless File Handling: Upload, process, and download files directly
-  - No Privacy Concerns: Fully isolated and secure execution
-
- 🔦 **Agents & Tools Integration**:  
-  - **[LibreChat Agents](https://www.librechat.ai/docs/features/agents)**:
-    - No-Code Custom Assistants: Build specialized, AI-driven helpers
-    - Agent Marketplace: Discover and deploy community-built agents
-    - Collaborative Sharing: Share agents with specific users and groups
-    - Flexible & Extensible: Use MCP Servers, tools, file search, code execution, and more
-    - Compatible with Custom Endpoints, OpenAI, Azure, Anthropic, AWS Bedrock, Google, Vertex AI, Responses API, and more
-    - [Model Context Protocol (MCP) Support](https://modelcontextprotocol.io/clients#librechat) for Tools
-
- 🔍 **Web Search**:  
-  - Search the internet and retrieve relevant information to enhance your AI context
-  - Combines search providers, content scrapers, and result rerankers for optimal results
-  - **Customizable Jina Reranking**: Configure custom Jina API URLs for reranking services
-  - **[Learn More →](https://www.librechat.ai/docs/features/web_search)**
-
- 🪄 **Generative UI with Code Artifacts**:  
-  - [Code Artifacts](https://youtu.be/GfTj7O4gmd0?si=WJbdnemZpJzBrJo3) allow creation of React, HTML, and Mermaid diagrams directly in chat
-
- 🎨 **Image Generation & Editing**
-  - Text-to-image and image-to-image with [GPT-Image-1](https://www.librechat.ai/docs/features/image_gen#1--openai-image-tools-recommended)
-  - Text-to-image with [DALL-E (3/2)](https://www.librechat.ai/docs/features/image_gen#2--dalle-legacy), [Stable Diffusion](https://www.librechat.ai/docs/features/image_gen#3--stable-diffusion-local), [Flux](https://www.librechat.ai/docs/features/image_gen#4--flux), or any [MCP server](https://www.librechat.ai/docs/features/image_gen#5--model-context-protocol-mcp)
-  - Produce stunning visuals from prompts or refine existing images with a single instruction
-
- 💾 **Presets & Context Management**:  
-  - Create, Save, & Share Custom Presets  
-  - Switch between AI Endpoints and Presets mid-chat
-  - Edit, Resubmit, and Continue Messages with Conversation branching  
-  - Create and share prompts with specific users and groups
-  - [Fork Messages & Conversations](https://www.librechat.ai/docs/features/fork) for Advanced Context control
-
- 💬 **Multimodal & File Interactions**:  
-  - Upload and analyze images with Claude 3, GPT-4.5, GPT-4o, o1, Llama-Vision, and Gemini 📸  
-  - Chat with Files using Custom Endpoints, OpenAI, Azure, Anthropic, AWS Bedrock, & Google 🗃️
-
- 🌎 **Multilingual UI**:
-  - English, 中文 (简体), 中文 (繁體), العربية, Deutsch, Español, Français, Italiano
-  - Polski, Português (PT), Português (BR), Русский, 日本語, Svenska, 한국어, Tiếng Việt
-  - Türkçe, Nederlands, עברית, Català, Čeština, Dansk, Eesti, فارسی
-  - Suomi, Magyar, Հայերեն, Bahasa Indonesia, ქართული, Latviešu, ไทย, ئۇيغۇرچە
-
- 🧠 **Reasoning UI**:  
-  - Dynamic Reasoning UI for Chain-of-Thought/Reasoning AI models like DeepSeek-R1
-
- 🎨 **Customizable Interface**:  
-  - Customizable Dropdown & Interface that adapts to both power users and newcomers
-
- 🗣️ **Speech & Audio**:  
-  - Chat hands-free with Speech-to-Text and Text-to-Speech  
-  - Automatically send and play Audio  
+- 🖥️ UI matching ChatGPT, including Dark mode, Streaming, and latest updates
+- 🤖 AI model selection:
+  - Anthropic (Claude), AWS Bedrock, OpenAI, Azure OpenAI, BingAI, ChatGPT, Google Vertex AI, Plugins, Assistants API (including Azure Assistants)
+- ✅ Compatible across both **[Remote & Local AI services](https://www.librechat.ai/docs/configuration/librechat_yaml/ai_endpoints):**
+  - groq, Ollama, Cohere, Mistral AI, Apple MLX, koboldcpp, OpenRouter, together.ai, Perplexity, ShuttleAI, and more
+- 🪄 Generative UI with **[Code Artifacts](https://youtu.be/GfTj7O4gmd0?si=WJbdnemZpJzBrJo3)**
+   - Create React, HTML code, and Mermaid diagrams right in chat
+- 💾 Create, Save, & Share Custom Presets
+- 🔀 Switch between AI Endpoints and Presets, mid-chat
+- 🔄 Edit, Resubmit, and Continue Messages with Conversation branching
+- 🌿 Fork Messages & Conversations for Advanced Context control
+- 💬 Multimodal Chat:
+    - Upload and analyze images with Claude 3, GPT-4 (including `gpt-4o` and `gpt-4o-mini`), and Gemini Vision 📸
+    - Chat with Files using Custom Endpoints, OpenAI, Azure, Anthropic, & Google. 🗃️
+    - Advanced Agents with Files, Code Interpreter, Tools, and API Actions 🔦
+      - Available through the [OpenAI Assistants API](https://platform.openai.com/docs/assistants/overview) 🌤️
+      - Non-OpenAI Agents in Active Development 🚧
+- 🌎 Multilingual UI:
+  - English, 中文, Deutsch, Español, Français, Italiano, Polski, Português Brasileiro,
+  - Русский, 日本語, Svenska, 한국어, Tiếng Việt, 繁體中文, العربية, Türkçe, Nederlands, עברית
+- 🎨 Customizable Dropdown & Interface: Adapts to both power users and newcomers
+- 📧 Verify your email to ensure secure access
+- 🗣️ Chat hands-free with Speech-to-Text and Text-to-Speech magic
+  - Automatically send and play Audio
  - Supports OpenAI, Azure OpenAI, and Elevenlabs
-
- 📥 **Import & Export Conversations**:  
-  - Import Conversations from LibreChat, ChatGPT, Chatbot UI  
-  - Export conversations as screenshots, markdown, text, json
-
- 🔍 **Search & Discovery**:  
-  - Search all messages/conversations
-
- 👥 **Multi-User & Secure Access**:
-  - Multi-User, Secure Authentication with OAuth2, LDAP, & Email Login Support
-  - Built-in Moderation, and Token spend tools
-
- ⚙️ **Configuration & Deployment**:  
-  - Configure Proxy, Reverse Proxy, Docker, & many Deployment options  
+- 📥 Import Conversations from LibreChat, ChatGPT, Chatbot UI
+- 📤 Export conversations as screenshots, markdown, text, json
+- 🔍 Search all messages/conversations
+- 🔌 Plugins, including web access, image generation with DALL-E-3 and more
+- 👥 Multi-User, Secure Authentication with Moderation and Token spend tools
+- ⚙️ Configure Proxy, Reverse Proxy, Docker, & many Deployment options:
  - Use completely local or deploy on the cloud
-
- 📖 **Open-Source & Community**:  
-  - Completely Open-Source & Built in Public  
-  - Community-driven development, support, and feedback
+- 📖 Completely Open-Source & Built in Public
+- 🧑‍🤝‍🧑 Community-driven development, support, and feedback

 [For a thorough review of our features, see our docs here](https://docs.librechat.ai/) 📚

@@ -141,8 +83,7 @@ LibreChat brings together the future of assistant AIs with the revolutionary tec

 With LibreChat, you no longer need to opt for ChatGPT Plus and can instead use free or pay-per-call APIs. We welcome contributions, cloning, and forking to enhance the capabilities of this advanced chatbot platform.

-[![Watch the video](https://raw.githubusercontent.com/LibreChat-AI/librechat.ai/main/public/images/changelog/v0.7.6.gif)](https://www.youtube.com/watch?v=ilfwGQtJNlI)
-
+[![Watch the video](https://raw.githubusercontent.com/LibreChat-AI/librechat.ai/main/public/images/changelog/v0.7.5.png)](https://www.youtube.com/watch?v=IDukQ7a2f3U)
 Click on the thumbnail to open the video☝️

 ---
@@ -155,8 +96,8 @@ Click on the thumbnail to open the video☝️

 **Other:**
  - **Website:** [librechat.ai](https://librechat.ai)
-  - **Documentation:** [librechat.ai/docs](https://librechat.ai/docs)
-  - **Blog:** [librechat.ai/blog](https://librechat.ai/blog)
+  - **Documentation:** [docs.librechat.ai](https://docs.librechat.ai)
+  - **Blog:** [blog.librechat.ai](https://blog.librechat.ai)

 ---

@@ -194,8 +135,6 @@ Contributions, suggestions, bug reports and fixes are welcome!

 For new features, components, or extensions, please open an issue and discuss before sending a PR.

-If you'd like to help translate LibreChat into your language, we'd love your contribution! Improving our translations not only makes LibreChat more accessible to users around the world but also enhances the overall user experience. Please check out our [Translation Guide](https://www.librechat.ai/docs/translation).
-
 ---

 ## 💖 This project exists in its current state thanks to all the people who contribute
@@ -203,15 +142,3 @@ If you'd like to help translate LibreChat into your language, we'd love your con
 <a href="https://github.com/danny-avila/LibreChat/graphs/contributors">
  <img src="https://contrib.rocks/image?repo=danny-avila/LibreChat" />
 </a>
-
---
-
-## 🎉 Special Thanks
-
-We thank [Locize](https://locize.com) for their translation management tools that support multiple languages in LibreChat.
-
-<p align="center">
-  <a href="https://locize.com" target="_blank" rel="noopener noreferrer">
-    <img src="https://github.com/user-attachments/assets/d6b70894-6064-475e-bb65-92a9e23e0077" alt="Locize Logo" height="50">
-  </a>
-</p>
--- a/api/app/bingai.js
+++ b/api/app/bingai.js
@@ -0,0 +1,112 @@
+require('dotenv').config();
+const { KeyvFile } = require('keyv-file');
+const { EModelEndpoint } = require('librechat-data-provider');
+const { getUserKey, checkUserKeyExpiry } = require('~/server/services/UserService');
+const { logger } = require('~/config');
+
+const askBing = async ({
+  text,
+  parentMessageId,
+  conversationId,
+  jailbreak,
+  jailbreakConversationId,
+  context,
+  systemMessage,
+  conversationSignature,
+  clientId,
+  invocationId,
+  toneStyle,
+  key: expiresAt,
+  onProgress,
+  userId,
+}) => {
+  const isUserProvided = process.env.BINGAI_TOKEN === 'user_provided';
+
+  let key = null;
+  if (expiresAt && isUserProvided) {
+    checkUserKeyExpiry(expiresAt, EModelEndpoint.bingAI);
+    key = await getUserKey({ userId, name: 'bingAI' });
+  }
+
+  const { BingAIClient } = await import('nodejs-gpt');
+  const store = {
+    store: new KeyvFile({ filename: './data/cache.json' }),
+  };
+
+  const bingAIClient = new BingAIClient({
+    // "_U" cookie from bing.com
+    // userToken:
+    //   isUserProvided ? key : process.env.BINGAI_TOKEN ?? null,
+    // If the above doesn't work, provide all your cookies as a string instead
+    cookies: isUserProvided ? key : process.env.BINGAI_TOKEN ?? null,
+    debug: false,
+    cache: store,
+    host: process.env.BINGAI_HOST || null,
+    proxy: process.env.PROXY || null,
+  });
+
+  let options = {};
+
+  if (jailbreakConversationId == 'false') {
+    jailbreakConversationId = false;
+  }
+
+  if (jailbreak) {
+    options = {
+      jailbreakConversationId: jailbreakConversationId || jailbreak,
+      context,
+      systemMessage,
+      parentMessageId,
+      toneStyle,
+      onProgress,
+      clientOptions: {
+        features: {
+          genImage: {
+            server: {
+              enable: true,
+              type: 'markdown_list',
+            },
+          },
+        },
+      },
+    };
+  } else {
+    options = {
+      conversationId,
+      context,
+      systemMessage,
+      parentMessageId,
+      toneStyle,
+      onProgress,
+      clientOptions: {
+        features: {
+          genImage: {
+            server: {
+              enable: true,
+              type: 'markdown_list',
+            },
+          },
+        },
+      },
+    };
+
+    // don't give those parameters for new conversation
+    // for new conversation, conversationSignature always is null
+    if (conversationSignature) {
+      options.encryptedConversationSignature = conversationSignature;
+      options.clientId = clientId;
+      options.invocationId = invocationId;
+    }
+  }
+
+  logger.debug('bing options', options);
+
+  const res = await bingAIClient.sendMessage(text, options);
+
+  return res;
+
+  // for reference:
+  // https://github.com/waylaidwanderer/node-chatgpt-api/blob/main/demos/use-bing-client.js
+};
+
+module.exports = { askBing };
--- a/api/app/chatgpt-browser.js
+++ b/api/app/chatgpt-browser.js
@@ -0,0 +1,57 @@
+require('dotenv').config();
+const { KeyvFile } = require('keyv-file');
+const { Constants, EModelEndpoint } = require('librechat-data-provider');
+const { getUserKey, checkUserKeyExpiry } = require('../server/services/UserService');
+
+const browserClient = async ({
+  text,
+  parentMessageId,
+  conversationId,
+  model,
+  key: expiresAt,
+  onProgress,
+  onEventMessage,
+  abortController,
+  userId,
+}) => {
+  const isUserProvided = process.env.CHATGPT_TOKEN === 'user_provided';
+
+  let key = null;
+  if (expiresAt && isUserProvided) {
+    checkUserKeyExpiry(expiresAt, EModelEndpoint.chatGPTBrowser);
+    key = await getUserKey({ userId, name: 'chatGPTBrowser' });
+  }
+
+  const { ChatGPTBrowserClient } = await import('nodejs-gpt');
+  const store = {
+    store: new KeyvFile({ filename: './data/cache.json' }),
+  };
+
+  const clientOptions = {
+    // Warning: This will expose your access token to a third party. Consider the risks before using this.
+    reverseProxyUrl:
+      process.env.CHATGPT_REVERSE_PROXY ?? 'https://ai.fakeopen.com/api/conversation',
+    // Access token from https://chat.openai.com/api/auth/session
+    accessToken: isUserProvided ? key : process.env.CHATGPT_TOKEN ?? null,
+    model: model,
+    debug: false,
+    proxy: process.env.PROXY ?? null,
+    user: userId,
+  };
+
+  const client = new ChatGPTBrowserClient(clientOptions, store);
+  let options = { onProgress, onEventMessage, abortController };
+
+  if (!!parentMessageId && !!conversationId) {
+    options = { ...options, parentMessageId, conversationId };
+  }
+
+  if (parentMessageId === Constants.NO_PARENT) {
+    delete options.conversationId;
+  }
+
+  const res = await client.sendMessage(text, options);
+  return res;
+};
+
+module.exports = { browserClient };
--- a/api/app/clients/AnthropicClient.js
+++ b/api/app/clients/AnthropicClient.js
@@ -1,26 +1,14 @@
 const Anthropic = require('@anthropic-ai/sdk');
 const { HttpsProxyAgent } = require('https-proxy-agent');
+const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
 const {
  Constants,
-  ErrorTypes,
  EModelEndpoint,
-  parseTextParts,
  anthropicSettings,
  getResponseSender,
  validateVisionModel,
 } = require('librechat-data-provider');
-const { SplitStreamHandler: _Handler } = require('@librechat/agents');
-const {
-  Tokenizer,
-  createFetch,
-  matchModelName,
-  getClaudeHeaders,
-  getModelMaxTokens,
-  configureReasoning,
-  checkPromptCacheSupport,
-  getModelMaxOutputTokens,
-  createStreamEventHandlers,
-} = require('@librechat/api');
+const { encodeAndFormat } = require('~/server/services/Files/images/encode');
 const {
  truncateText,
  formatMessage,
@@ -29,8 +17,8 @@ const {
  parseParamFromPrompt,
  createContextHandlers,
 } = require('./prompts');
+const { getModelMaxTokens, getModelMaxOutputTokens, matchModelName } = require('~/utils');
 const { spendTokens, spendStructuredTokens } = require('~/models/spendTokens');
-const { encodeAndFormat } = require('~/server/services/Files/images/encode');
 const { sleep } = require('~/server/utils');
 const BaseClient = require('./BaseClient');
 const { logger } = require('~/config');
@@ -38,14 +26,7 @@ const { logger } = require('~/config');
 const HUMAN_PROMPT = '\n\nHuman:';
 const AI_PROMPT = '\n\nAssistant:';

-class SplitStreamHandler extends _Handler {
-  getDeltaContent(chunk) {
-    return (chunk?.delta?.text ?? chunk?.completion) || '';
-  }
-  getReasoningDelta(chunk) {
-    return chunk?.delta?.thinking || '';
-  }
-}
+const tokenizersCache = {};

 /** Helper function to introduce a delay before retrying */
 function delayBeforeRetry(attempts, baseDelay = 1000) {
@@ -73,10 +54,13 @@ class AnthropicClient extends BaseClient {
    this.message_delta;
    /** Whether the model is part of the Claude 3 Family
     * @type {boolean} */
-    this.isClaudeLatest;
+    this.isClaude3;
    /** Whether to use Messages API or Completions API
     * @type {boolean} */
    this.useMessages;
+    /** Whether or not the model is limited to the legacy amount of output tokens
+     * @type {boolean} */
+    this.isLegacyOutput;
    /** Whether or not the model supports Prompt Caching
     * @type {boolean} */
    this.supportsCacheControl;
@@ -86,8 +70,6 @@ class AnthropicClient extends BaseClient {
    /** The key for the usage object's output tokens
     * @type {string} */
    this.outputTokensKey = 'output_tokens';
-    /** @type {SplitStreamHandler | undefined} */
-    this.streamHandler;
  }

  setOptions(options) {
@@ -116,25 +98,20 @@ class AnthropicClient extends BaseClient {
    );

    const modelMatch = matchModelName(this.modelOptions.model, EModelEndpoint.anthropic);
-    this.isClaudeLatest =
-      /claude-[3-9]/.test(modelMatch) || /claude-(?:sonnet|opus|haiku)-[4-9]/.test(modelMatch);
-    const isLegacyOutput = !(
-      /claude-3[-.]5-sonnet/.test(modelMatch) ||
-      /claude-3[-.]7/.test(modelMatch) ||
-      /claude-(?:sonnet|opus|haiku)-[4-9]/.test(modelMatch) ||
-      /claude-[4-9]/.test(modelMatch)
-    );
-    this.supportsCacheControl = this.options.promptCache && checkPromptCacheSupport(modelMatch);
+    this.isClaude3 = modelMatch.includes('claude-3');
+    this.isLegacyOutput = !modelMatch.includes('claude-3-5-sonnet');
+    this.supportsCacheControl =
+      this.options.promptCache && this.checkPromptCacheSupport(modelMatch);

    if (
-      isLegacyOutput &&
+      this.isLegacyOutput &&
      this.modelOptions.maxOutputTokens &&
      this.modelOptions.maxOutputTokens > legacy.maxOutputTokens.default
    ) {
      this.modelOptions.maxOutputTokens = legacy.maxOutputTokens.default;
    }

-    this.useMessages = this.isClaudeLatest || !!this.options.attachments;
+    this.useMessages = this.isClaude3 || !!this.options.attachments;

    this.defaultVisionModel = this.options.visionModel ?? 'claude-3-sonnet-20240229';
    this.options.attachments?.then((attachments) => this.checkVisionRequest(attachments));
@@ -150,21 +127,16 @@ class AnthropicClient extends BaseClient {
        this.options.endpointType ?? this.options.endpoint,
        this.options.endpointTokenConfig,
      ) ??
-      anthropicSettings.maxOutputTokens.reset(this.modelOptions.model);
+      1500;
    this.maxPromptTokens =
      this.options.maxPromptTokens || this.maxContextTokens - this.maxResponseTokens;

-    const reservedTokens = this.maxPromptTokens + this.maxResponseTokens;
-    if (reservedTokens > this.maxContextTokens) {
-      const info = `Total Possible Tokens + Max Output Tokens must be less than or equal to Max Context Tokens: ${this.maxPromptTokens} (total possible output) + ${this.maxResponseTokens} (max output) = ${reservedTokens}/${this.maxContextTokens} (max context)`;
-      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
-      logger.warn(info);
-      throw new Error(errorMessage);
-    } else if (this.maxResponseTokens === this.maxContextTokens) {
-      const info = `Max Output Tokens must be less than Max Context Tokens: ${this.maxResponseTokens} (max output) = ${this.maxContextTokens} (max context)`;
-      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
-      logger.warn(info);
-      throw new Error(errorMessage);
+    if (this.maxPromptTokens + this.maxResponseTokens > this.maxContextTokens) {
+      throw new Error(
+        `maxPromptTokens + maxOutputTokens (${this.maxPromptTokens} + ${this.maxResponseTokens} = ${
+          this.maxPromptTokens + this.maxResponseTokens
+        }) must be less than or equal to maxContextTokens (${this.maxContextTokens})`,
+      );
    }

    this.sender =
@@ -177,6 +149,7 @@ class AnthropicClient extends BaseClient {

    this.startToken = '||>';
    this.endToken = '';
+    this.gptEncoder = this.constructor.getTokenizer('cl100k_base');

    return this;
  }
@@ -189,25 +162,30 @@ class AnthropicClient extends BaseClient {
  getClient(requestOptions) {
    /** @type {Anthropic.ClientOptions} */
    const options = {
-      fetch: createFetch({
-        directEndpoint: this.options.directEndpoint,
-        reverseProxyUrl: this.options.reverseProxyUrl,
-      }),
+      fetch: this.fetch,
      apiKey: this.apiKey,
-      fetchOptions: {},
    };

    if (this.options.proxy) {
-      options.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
+      options.httpAgent = new HttpsProxyAgent(this.options.proxy);
    }

    if (this.options.reverseProxyUrl) {
      options.baseURL = this.options.reverseProxyUrl;
    }

-    const headers = getClaudeHeaders(requestOptions?.model, this.supportsCacheControl);
-    if (headers) {
-      options.defaultHeaders = headers;
+    if (
+      this.supportsCacheControl &&
+      requestOptions?.model &&
+      requestOptions.model.includes('claude-3-5-sonnet')
+    ) {
+      options.defaultHeaders = {
+        'anthropic-beta': 'max-tokens-3-5-sonnet-2024-07-15,prompt-caching-2024-07-31',
+      };
+    } else if (this.supportsCacheControl) {
+      options.defaultHeaders = {
+        'anthropic-beta': 'prompt-caching-2024-07-31',
+      };
    }

    return new Anthropic(options);
@@ -401,13 +379,13 @@ class AnthropicClient extends BaseClient {
    const formattedMessages = orderedMessages.map((message, i) => {
      const formattedMessage = this.useMessages
        ? formatMessage({
-            message,
-            endpoint: EModelEndpoint.anthropic,
-          })
+          message,
+          endpoint: EModelEndpoint.anthropic,
+        })
        : {
-            author: message.isCreatedByUser ? this.userLabel : this.assistantLabel,
-            content: message?.content ?? message.text,
-          };
+          author: message.isCreatedByUser ? this.userLabel : this.assistantLabel,
+          content: message?.content ?? message.text,
+        };

      const needsTokenCount = this.contextStrategy && !orderedMessages[i].tokenCount;
      /* If tokens were never counted, or, is a Vision request and the message has files, count again */
@@ -423,9 +401,6 @@ class AnthropicClient extends BaseClient {
            this.contextHandlers?.processFile(file);
            continue;
          }
-          if (file.metadata?.fileIdentifier) {
-            continue;
-          }

          orderedMessages[i].tokenCount += this.calculateImageTokenCost({
            width: file.width,
@@ -444,7 +419,7 @@ class AnthropicClient extends BaseClient {
    }

    let { context: messagesInWindow, remainingContextTokens } =
-      await this.getMessagesWithinTokenLimit({ messages: formattedMessages });
+      await this.getMessagesWithinTokenLimit(formattedMessages);

    const tokenCountMap = orderedMessages
      .slice(orderedMessages.length - messagesInWindow.length)
@@ -659,10 +634,7 @@ class AnthropicClient extends BaseClient {
      );
    };

-    if (
-      /claude-[3-9]/.test(this.modelOptions.model) ||
-      /claude-(?:sonnet|opus|haiku)-[4-9]/.test(this.modelOptions.model)
-    ) {
+    if (this.modelOptions.model.includes('claude-3')) {
      await buildMessagesPayload();
      processTokens();
      return {
@@ -688,7 +660,7 @@ class AnthropicClient extends BaseClient {
  }

  getCompletion() {
-    logger.debug("AnthropicClient doesn't use getCompletion (all handled in sendCompletion)");
+    logger.debug('AnthropicClient doesn\'t use getCompletion (all handled in sendCompletion)');
  }

  /**
@@ -699,41 +671,29 @@ class AnthropicClient extends BaseClient {
   * @returns {Promise<Anthropic.default.Message | Anthropic.default.Completion>} The response from the Anthropic client.
   */
  async createResponse(client, options, useMessages) {
-    return (useMessages ?? this.useMessages)
+    return useMessages ?? this.useMessages
      ? await client.messages.create(options)
      : await client.completions.create(options);
  }

-  getMessageMapMethod() {
-    /**
-     * @param {TMessage} msg
-     */
-    return (msg) => {
-      if (msg.text != null && msg.text && msg.text.startsWith(':::thinking')) {
-        msg.text = msg.text.replace(/:::thinking.*?:::/gs, '').trim();
-      } else if (msg.content != null) {
-        msg.text = parseTextParts(msg.content, true);
-        delete msg.content;
-      }
-
-      return msg;
-    };
-  }
-
  /**
-   * @param {string[]} [intermediateReply]
-   * @returns {string}
+   * @param {string} modelName
+   * @returns {boolean}
   */
-  getStreamText(intermediateReply) {
-    if (!this.streamHandler) {
-      return intermediateReply?.join('') ?? '';
+  checkPromptCacheSupport(modelName) {
+    const modelMatch = matchModelName(modelName, EModelEndpoint.anthropic);
+    if (modelMatch.includes('claude-3-5-sonnet-latest')) {
+      return false;
    }
-
-    const reasoningText = this.streamHandler.reasoningTokens.join('');
-
-    const reasoningBlock = reasoningText.length > 0 ? `:::thinking\n${reasoningText}\n:::\n` : '';
-
-    return `${reasoningBlock}${this.streamHandler.tokens.join('')}`;
+    if (
+      modelMatch === 'claude-3-5-sonnet' ||
+      modelMatch === 'claude-3-5-haiku' ||
+      modelMatch === 'claude-3-haiku' ||
+      modelMatch === 'claude-3-opus'
+    ) {
+      return true;
+    }
+    return false;
  }

  async sendCompletion(payload, { onProgress, abortController }) {
@@ -753,6 +713,7 @@ class AnthropicClient extends BaseClient {
      user_id: this.user,
    };

+    let text = '';
    const {
      stream,
      model,
@@ -763,34 +724,22 @@ class AnthropicClient extends BaseClient {
      topK: top_k,
    } = this.modelOptions;

-    let requestOptions = {
+    const requestOptions = {
      model,
      stream: stream || true,
      stop_sequences,
      temperature,
      metadata,
+      top_p,
+      top_k,
    };

    if (this.useMessages) {
      requestOptions.messages = payload;
-      requestOptions.max_tokens =
-        maxOutputTokens || anthropicSettings.maxOutputTokens.reset(requestOptions.model);
+      requestOptions.max_tokens = maxOutputTokens || legacy.maxOutputTokens.default;
    } else {
      requestOptions.prompt = payload;
-      requestOptions.max_tokens_to_sample = maxOutputTokens || legacy.maxOutputTokens.default;
-    }
-
-    requestOptions = configureReasoning(requestOptions, {
-      thinking: this.options.thinking,
-      thinkingBudget: this.options.thinkingBudget,
-    });
-
-    if (!/claude-3[-.]7/.test(model)) {
-      requestOptions.top_p = top_p;
-      requestOptions.top_k = top_k;
-    } else if (requestOptions.thinking == null) {
-      requestOptions.topP = top_p;
-      requestOptions.topK = top_k;
+      requestOptions.max_tokens_to_sample = maxOutputTokens || 1500;
    }

    if (this.systemMessage && this.supportsCacheControl === true) {
@@ -810,14 +759,13 @@ class AnthropicClient extends BaseClient {
    }

    logger.debug('[AnthropicClient]', { ...requestOptions });
-    const handlers = createStreamEventHandlers(this.options.res);
-    this.streamHandler = new SplitStreamHandler({
-      accumulate: true,
-      runId: this.responseMessageId,
-      handlers,
-    });

-    let intermediateReply = this.streamHandler.tokens;
+    const handleChunk = (currentChunk) => {
+      if (currentChunk) {
+        text += currentChunk;
+        onProgress(currentChunk);
+      }
+    };

    const maxRetries = 3;
    const streamRate = this.options.streamRate ?? Constants.DEFAULT_STREAM_RATE;
@@ -838,15 +786,22 @@ class AnthropicClient extends BaseClient {
          });

          for await (const completion of response) {
+            // Handle each completion as before
            const type = completion?.type ?? '';
            if (tokenEventTypes.has(type)) {
              logger.debug(`[AnthropicClient] ${type}`, completion);
              this[type] = completion;
            }
-            this.streamHandler.handle(completion);
+            if (completion?.delta?.text) {
+              handleChunk(completion.delta.text);
+            } else if (completion.completion) {
+              handleChunk(completion.completion);
+            }
+
            await sleep(streamRate);
          }

+          // Successful processing, exit loop
          break;
        } catch (error) {
          attempts += 1;
@@ -856,10 +811,6 @@ class AnthropicClient extends BaseClient {

          if (attempts < maxRetries) {
            await delayBeforeRetry(attempts, 350);
-          } else if (this.streamHandler && this.streamHandler.reasoningTokens.length) {
-            return this.getStreamText();
-          } else if (intermediateReply.length > 0) {
-            return this.getStreamText(intermediateReply);
          } else {
            throw new Error(`Operation failed after ${maxRetries} attempts: ${error.message}`);
          }
@@ -875,7 +826,8 @@ class AnthropicClient extends BaseClient {
    }

    await processResponse.bind(this)();
-    return this.getStreamText(intermediateReply);
+
+    return text.trim();
  }

  getSaveOptions() {
@@ -885,8 +837,6 @@ class AnthropicClient extends BaseClient {
      promptPrefix: this.options.promptPrefix,
      modelLabel: this.options.modelLabel,
      promptCache: this.options.promptCache,
-      thinking: this.options.thinking,
-      thinkingBudget: this.options.thinkingBudget,
      resendFiles: this.options.resendFiles,
      iconURL: this.options.iconURL,
      greeting: this.options.greeting,
@@ -896,21 +846,25 @@ class AnthropicClient extends BaseClient {
  }

  getBuildMessagesOptions() {
-    logger.debug("AnthropicClient doesn't use getBuildMessagesOptions");
+    logger.debug('AnthropicClient doesn\'t use getBuildMessagesOptions');
  }

-  getEncoding() {
-    return 'cl100k_base';
+  static getTokenizer(encoding, isModelName = false, extendSpecialTokens = {}) {
+    if (tokenizersCache[encoding]) {
+      return tokenizersCache[encoding];
+    }
+    let tokenizer;
+    if (isModelName) {
+      tokenizer = encodingForModel(encoding, extendSpecialTokens);
+    } else {
+      tokenizer = getEncoding(encoding, extendSpecialTokens);
+    }
+    tokenizersCache[encoding] = tokenizer;
+    return tokenizer;
  }

-  /**
-   * Returns the token count of a given text. It also checks and resets the tokenizers if necessary.
-   * @param {string} text - The text to get the token count for.
-   * @returns {number} The token count of the given text.
-   */
  getTokenCount(text) {
-    const encoding = this.getEncoding();
-    return Tokenizer.getTokenCount(text, encoding);
+    return this.gptEncoder.encode(text, 'all').length;
  }

  /**
--- a/api/app/clients/BaseClient.js
+++ b/api/app/clients/BaseClient.js
@@ -1,22 +1,21 @@
 const crypto = require('crypto');
 const fetch = require('node-fetch');
-const { logger } = require('@librechat/data-schemas');
-const { getBalanceConfig } = require('@librechat/api');
 const {
  supportsBalanceCheck,
  isAgentsEndpoint,
  isParamEndpoint,
-  EModelEndpoint,
-  ContentTypes,
-  excludedKeys,
  ErrorTypes,
  Constants,
+  CacheKeys,
+  Time,
 } = require('librechat-data-provider');
-const { getMessages, saveMessage, updateMessage, saveConvo, getConvo } = require('~/models');
-const { checkBalance } = require('~/models/balanceMethods');
-const { truncateToolCallOutputs } = require('./prompts');
+const { getMessages, saveMessage, updateMessage, saveConvo } = require('~/models');
+const { addSpaceIfNeeded, isEnabled } = require('~/server/utils');
+const checkBalance = require('~/models/checkBalance');
 const { getFiles } = require('~/models/File');
+const { getLogStores } = require('~/cache');
 const TextStream = require('./TextStream');
+const { logger } = require('~/config');

 class BaseClient {
  constructor(apiKey, options = {}) {
@@ -28,18 +27,21 @@ class BaseClient {
      month: 'long',
      day: 'numeric',
    });
+    this.fetch = this.fetch.bind(this);
    /** @type {boolean} */
    this.skipSaveConvo = false;
    /** @type {boolean} */
    this.skipSaveUserMessage = false;
+    /** @type {ClientDatabaseSavePromise} */
+    this.userMessagePromise;
+    /** @type {ClientDatabaseSavePromise} */
+    this.responsePromise;
    /** @type {string} */
    this.user;
    /** @type {string} */
    this.conversationId;
    /** @type {string} */
    this.responseMessageId;
-    /** @type {string} */
-    this.parentMessageId;
    /** @type {TAttachment[]} */
    this.attachments;
    /** The key for the usage object's input tokens
@@ -48,32 +50,18 @@ class BaseClient {
    /** The key for the usage object's output tokens
     * @type {string} */
    this.outputTokensKey = 'completion_tokens';
-    /** @type {Set<string>} */
-    this.savedMessageIds = new Set();
-    /**
-     * Flag to determine if the client re-submitted the latest assistant message.
-     * @type {boolean | undefined} */
-    this.continued;
-    /**
-     * Flag to determine if the client has already fetched the conversation while saving new messages.
-     * @type {boolean | undefined} */
-    this.fetchedConvo;
-    /** @type {TMessage[]} */
-    this.currentMessages = [];
-    /** @type {import('librechat-data-provider').VisionModes | undefined} */
-    this.visionMode;
  }

  setOptions() {
-    throw new Error("Method 'setOptions' must be implemented.");
+    throw new Error('Method \'setOptions\' must be implemented.');
  }

  async getCompletion() {
-    throw new Error("Method 'getCompletion' must be implemented.");
+    throw new Error('Method \'getCompletion\' must be implemented.');
  }

  async sendCompletion() {
-    throw new Error("Method 'sendCompletion' must be implemented.");
+    throw new Error('Method \'sendCompletion\' must be implemented.');
  }

  getSaveOptions() {
@@ -96,7 +84,7 @@ class BaseClient {
      return this.options.agent.id;
    }

-    return this.modelOptions?.model ?? this.model;
+    return this.modelOptions.model;
  }

  /**
@@ -105,23 +93,18 @@ class BaseClient {
   * @returns {number}
   */
  getTokenCountForResponse(responseMessage) {
-    logger.debug('[BaseClient] `recordTokenUsage` not implemented.', responseMessage);
+    logger.debug('`[BaseClient] recordTokenUsage` not implemented.', responseMessage);
  }

  /**
   * Abstract method to record token usage. Subclasses must implement this method.
   * If a correction to the token usage is needed, the method should return an object with the corrected token counts.
-   * Should only be used if `recordCollectedUsage` was not used instead.
-   * @param {string} [model]
-   * @param {AppConfig['balance']} [balance]
   * @param {number} promptTokens
   * @param {number} completionTokens
   * @returns {Promise<void>}
   */
-  async recordTokenUsage({ model, balance, promptTokens, completionTokens }) {
-    logger.debug('[BaseClient] `recordTokenUsage` not implemented.', {
-      model,
-      balance,
+  async recordTokenUsage({ promptTokens, completionTokens }) {
+    logger.debug('`[BaseClient] recordTokenUsage` not implemented.', {
      promptTokens,
      completionTokens,
    });
@@ -190,8 +173,7 @@ class BaseClient {
    this.user = user;
    const saveOptions = this.getSaveOptions();
    this.abortController = opts.abortController ?? new AbortController();
-    const requestConvoId = overrideConvoId ?? opts.conversationId;
-    const conversationId = requestConvoId ?? crypto.randomUUID();
+    const conversationId = overrideConvoId ?? opts.conversationId ?? crypto.randomUUID();
    const parentMessageId = opts.parentMessageId ?? Constants.NO_PARENT;
    const userMessageId =
      overrideUserMessageId ?? opts.overrideParentMessageId ?? crypto.randomUUID();
@@ -206,22 +188,17 @@ class BaseClient {
      this.currentMessages[this.currentMessages.length - 1].messageId = head;
    }

-    if (opts.isRegenerate && responseMessageId.endsWith('_')) {
-      responseMessageId = crypto.randomUUID();
-    }
-
    this.responseMessageId = responseMessageId;

    return {
      ...opts,
      user,
      head,
-      saveOptions,
-      userMessageId,
-      requestConvoId,
      conversationId,
      parentMessageId,
+      userMessageId,
      responseMessageId,
+      saveOptions,
    };
  }

@@ -233,7 +210,6 @@ class BaseClient {
      sender: 'User',
      text,
      isCreatedByUser: true,
-      targetModel: this.modelOptions?.model ?? this.model,
    };
  }

@@ -241,22 +217,21 @@ class BaseClient {
    const {
      user,
      head,
-      saveOptions,
-      userMessageId,
-      requestConvoId,
      conversationId,
      parentMessageId,
+      userMessageId,
      responseMessageId,
+      saveOptions,
    } = await this.setMessageOptions(opts);

    const userMessage = opts.isEdited
      ? this.currentMessages[this.currentMessages.length - 2]
      : this.createUserMessage({
-          messageId: userMessageId,
-          parentMessageId,
-          conversationId,
-          text: message,
-        });
+        messageId: userMessageId,
+        parentMessageId,
+        conversationId,
+        text: message,
+      });

    if (typeof opts?.getReqData === 'function') {
      opts.getReqData({
@@ -268,8 +243,7 @@ class BaseClient {
    }

    if (typeof opts?.onStart === 'function') {
-      const isNewConvo = !requestConvoId && parentMessageId === Constants.NO_PARENT;
-      opts.onStart(userMessage, responseMessageId, isNewConvo);
+      opts.onStart(userMessage, responseMessageId);
    }

    return {
@@ -286,24 +260,17 @@ class BaseClient {
  /**
   * Adds instructions to the messages array. If the instructions object is empty or undefined,
   * the original messages array is returned. Otherwise, the instructions are added to the messages
-   * array either at the beginning (default) or preserving the last message at the end.
+   * array, preserving the last message at the end.
   *
   * @param {Array} messages - An array of messages.
   * @param {Object} instructions - An object containing instructions to be added to the messages.
-   * @param {boolean} [beforeLast=false] - If true, adds instructions before the last message; if false, adds at the beginning.
   * @returns {Array} An array containing messages and instructions, or the original messages if instructions are empty.
   */
-  addInstructions(messages, instructions, beforeLast = false) {
+  addInstructions(messages, instructions) {
+    const payload = [];
    if (!instructions || Object.keys(instructions).length === 0) {
      return messages;
    }
-
-    if (!beforeLast) {
-      return [instructions, ...messages];
-    }
-
-    // Legacy behavior: add instructions before the last message
-    const payload = [];
    if (messages.length > 1) {
      payload.push(...messages.slice(0, -1));
    }
@@ -318,9 +285,6 @@ class BaseClient {
  }

  async handleTokenCountMap(tokenCountMap) {
-    if (this.clientName === EModelEndpoint.agents) {
-      return;
-    }
    if (this.currentMessages.length === 0) {
      return;
    }
@@ -369,35 +333,25 @@ class BaseClient {
   * If the token limit would be exceeded by adding a message, that message is not added to the context and remains in the original array.
   * The method uses `push` and `pop` operations for efficient array manipulation, and reverses the context array at the end to maintain the original order of the messages.
   *
-   * @param {Object} params
-   * @param {TMessage[]} params.messages - An array of messages, each with a `tokenCount` property. The messages should be ordered from oldest to newest.
-   * @param {number} [params.maxContextTokens] - The max number of tokens allowed in the context. If not provided, defaults to `this.maxContextTokens`.
-   * @param {{ role: 'system', content: text, tokenCount: number }} [params.instructions] - Instructions already added to the context at index 0.
-   * @returns {Promise<{
-   *  context: TMessage[],
-   *  remainingContextTokens: number,
-   *  messagesToRefine: TMessage[],
-   * }>} An object with three properties: `context`, `remainingContextTokens`, and `messagesToRefine`.
+   * @param {Array} _messages - An array of messages, each with a `tokenCount` property. The messages should be ordered from oldest to newest.
+   * @param {number} [maxContextTokens] - The max number of tokens allowed in the context. If not provided, defaults to `this.maxContextTokens`.
+   * @returns {Object} An object with four properties: `context`, `summaryIndex`, `remainingContextTokens`, and `messagesToRefine`.
   *    `context` is an array of messages that fit within the token limit.
+   *    `summaryIndex` is the index of the first message in the `messagesToRefine` array.
   *    `remainingContextTokens` is the number of tokens remaining within the limit after adding the messages to the context.
   *    `messagesToRefine` is an array of messages that were not added to the context because they would have exceeded the token limit.
   */
-  async getMessagesWithinTokenLimit({ messages: _messages, maxContextTokens, instructions }) {
+  async getMessagesWithinTokenLimit(_messages, maxContextTokens) {
    // Every reply is primed with <|start|>assistant<|message|>, so we
    // start with 3 tokens for the label after all messages have been counted.
    let currentTokenCount = 3;
-    const instructionsTokenCount = instructions?.tokenCount ?? 0;
-    let remainingContextTokens =
-      (maxContextTokens ?? this.maxContextTokens) - instructionsTokenCount;
+    let summaryIndex = -1;
+    let remainingContextTokens = maxContextTokens ?? this.maxContextTokens;
    const messages = [..._messages];

    const context = [];
-
    if (currentTokenCount < remainingContextTokens) {
      while (messages.length > 0 && currentTokenCount < remainingContextTokens) {
-        if (messages.length === 1 && instructions) {
-          break;
-        }
        const poppedMessage = messages.pop();
        const { tokenCount } = poppedMessage;

@@ -411,18 +365,15 @@ class BaseClient {
      }
    }

-    if (instructions) {
-      context.push(_messages[0]);
-      messages.shift();
-    }
-
    const prunedMemory = messages;
+    summaryIndex = prunedMemory.length - 1;
    remainingContextTokens -= currentTokenCount;

    return {
      context: context.reverse(),
      remainingContextTokens,
      messagesToRefine: prunedMemory,
+      summaryIndex,
    };
  }

@@ -438,38 +389,12 @@ class BaseClient {
    if (instructions) {
      ({ tokenCount, ..._instructions } = instructions);
    }
-
    _instructions && logger.debug('[BaseClient] instructions tokenCount: ' + tokenCount);
-    if (tokenCount && tokenCount > this.maxContextTokens) {
-      const info = `${tokenCount} / ${this.maxContextTokens}`;
-      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
-      logger.warn(`Instructions token count exceeds max token count (${info}).`);
-      throw new Error(errorMessage);
-    }
-
-    if (this.clientName === EModelEndpoint.agents) {
-      const { dbMessages, editedIndices } = truncateToolCallOutputs(
-        orderedMessages,
-        this.maxContextTokens,
-        this.getTokenCountForMessage.bind(this),
-      );
-
-      if (editedIndices.length > 0) {
-        logger.debug('[BaseClient] Truncated tool call outputs:', editedIndices);
-        for (const index of editedIndices) {
-          formattedMessages[index].content = dbMessages[index].content;
-        }
-        orderedMessages = dbMessages;
-      }
-    }
-
+    let payload = this.addInstructions(formattedMessages, _instructions);
    let orderedWithInstructions = this.addInstructions(orderedMessages, instructions);

-    let { context, remainingContextTokens, messagesToRefine } =
-      await this.getMessagesWithinTokenLimit({
-        messages: orderedWithInstructions,
-        instructions,
-      });
+    let { context, remainingContextTokens, messagesToRefine, summaryIndex } =
+      await this.getMessagesWithinTokenLimit(orderedWithInstructions);

    logger.debug('[BaseClient] Context Count (1/2)', {
      remainingContextTokens,
@@ -481,9 +406,7 @@ class BaseClient {
    let { shouldSummarize } = this;

    // Calculate the difference in length to determine how many messages were discarded if any
-    let payload;
-    let { length } = formattedMessages;
-    length += instructions != null ? 1 : 0;
+    const { length } = payload;
    const diff = length - context.length;
    const firstMessage = orderedWithInstructions[0];
    const usePrevSummary =
@@ -493,31 +416,18 @@ class BaseClient {
      this.previous_summary.messageId === firstMessage.messageId;

    if (diff > 0) {
-      payload = formattedMessages.slice(diff);
+      payload = payload.slice(diff);
      logger.debug(
        `[BaseClient] Difference between original payload (${length}) and context (${context.length}): ${diff}`,
      );
    }

-    payload = this.addInstructions(payload ?? formattedMessages, _instructions);
-
    const latestMessage = orderedWithInstructions[orderedWithInstructions.length - 1];
    if (payload.length === 0 && !shouldSummarize && latestMessage) {
      const info = `${latestMessage.tokenCount} / ${this.maxContextTokens}`;
      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
      logger.warn(`Prompt token count exceeds max token count (${info}).`);
      throw new Error(errorMessage);
-    } else if (
-      _instructions &&
-      payload.length === 1 &&
-      payload[0].content === _instructions.content
-    ) {
-      const info = `${tokenCount + 3} / ${this.maxContextTokens}`;
-      const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
-      logger.warn(
-        `Including instructions, the prompt token count exceeds remaining max token count (${info}).`,
-      );
-      throw new Error(errorMessage);
    }

    if (usePrevSummary) {
@@ -535,7 +445,7 @@ class BaseClient {
    }

    // Make sure to only continue summarization logic if the summary message was generated
-    shouldSummarize = summaryMessage != null && shouldSummarize === true;
+    shouldSummarize = summaryMessage && shouldSummarize;

    logger.debug('[BaseClient] Context Count (2/2)', {
      remainingContextTokens,
@@ -545,18 +455,17 @@ class BaseClient {
    /** @type {Record<string, number> | undefined} */
    let tokenCountMap;
    if (buildTokenMap) {
-      const currentPayload = shouldSummarize ? orderedWithInstructions : context;
-      tokenCountMap = currentPayload.reduce((map, message, index) => {
+      tokenCountMap = orderedWithInstructions.reduce((map, message, index) => {
        const { messageId } = message;
        if (!messageId) {
          return map;
        }

-        if (shouldSummarize && index === messagesToRefine.length - 1 && !usePrevSummary) {
+        if (shouldSummarize && index === summaryIndex && !usePrevSummary) {
          map.summaryMessage = { ...summaryMessage, messageId, tokenCount: summaryTokenCount };
        }

-        map[messageId] = currentPayload[index].tokenCount;
+        map[messageId] = orderedWithInstructions[index].tokenCount;
        return map;
      }, {});
    }
@@ -575,9 +484,6 @@ class BaseClient {
  }

  async sendMessage(message, opts = {}) {
-    const appConfig = this.options.req?.config;
-    /** @type {Promise<TMessage>} */
-    let userMessagePromise;
    const { user, head, isEdited, conversationId, responseMessageId, saveOptions, userMessage } =
      await this.handleStartMethods(message, opts);

@@ -589,7 +495,7 @@ class BaseClient {
      });
    }

-    const { editedContent } = opts;
+    const { generation = '' } = opts;

    // It's not necessary to push to currentMessages
    // depending on subclass implementation of handling messages
@@ -602,42 +508,27 @@ class BaseClient {
          conversationId,
          parentMessageId: userMessage.messageId,
          isCreatedByUser: false,
-          model: this.modelOptions?.model ?? this.model,
+          model: this.modelOptions.model,
          sender: this.sender,
+          text: generation,
        };
        this.currentMessages.push(userMessage, latestMessage);
-      } else if (editedContent != null) {
-        // Handle editedContent for content parts
-        if (editedContent && latestMessage.content && Array.isArray(latestMessage.content)) {
-          const { index, text, type } = editedContent;
-          if (index >= 0 && index < latestMessage.content.length) {
-            const contentPart = latestMessage.content[index];
-            if (type === ContentTypes.THINK && contentPart.type === ContentTypes.THINK) {
-              contentPart[ContentTypes.THINK] = text;
-            } else if (type === ContentTypes.TEXT && contentPart.type === ContentTypes.TEXT) {
-              contentPart[ContentTypes.TEXT] = text;
-            }
-          }
-        }
+      } else {
+        latestMessage.text = generation;
      }
-      this.continued = true;
    } else {
      this.currentMessages.push(userMessage);
    }

-    /**
-     * When the userMessage is pushed to currentMessages, the parentMessage is the userMessageId.
-     * this only matters when buildMessages is utilizing the parentMessageId, and may vary on implementation
-     */
-    const parentMessageId = isEdited ? head : userMessage.messageId;
-    this.parentMessageId = parentMessageId;
    let {
      prompt: payload,
      tokenCountMap,
      promptTokens,
    } = await this.buildMessages(
      this.currentMessages,
-      parentMessageId,
+      // When the userMessage is pushed to currentMessages, the parentMessage is the userMessageId.
+      // this only matters when buildMessages is utilizing the parentMessageId, and may vary on implementation
+      isEdited ? head : userMessage.messageId,
      this.getBuildMessagesOptions(opts),
      opts,
    );
@@ -653,18 +544,16 @@ class BaseClient {
    }

    if (!isEdited && !this.skipSaveUserMessage) {
-      userMessagePromise = this.saveMessageToDatabase(userMessage, saveOptions, user);
-      this.savedMessageIds.add(userMessage.messageId);
+      this.userMessagePromise = this.saveMessageToDatabase(userMessage, saveOptions, user);
      if (typeof opts?.getReqData === 'function') {
        opts.getReqData({
-          userMessagePromise,
+          userMessagePromise: this.userMessagePromise,
        });
      }
    }

-    const balanceConfig = getBalanceConfig(appConfig);
    if (
-      balanceConfig?.enabled &&
+      isEnabled(process.env.CHECK_BALANCE) &&
      supportsBalanceCheck[this.options.endpointType ?? this.options.endpoint]
    ) {
      await checkBalance({
@@ -674,8 +563,8 @@ class BaseClient {
          user: this.user,
          tokenType: 'prompt',
          amount: promptTokens,
+          model: this.modelOptions.model,
          endpoint: this.options.endpoint,
-          model: this.modelOptions?.model ?? this.model,
          endpointTokenConfig: this.options.endpointTokenConfig,
        },
      });
@@ -683,11 +572,8 @@ class BaseClient {

    /** @type {string|string[]|undefined} */
    const completion = await this.sendCompletion(payload, opts);
-    if (this.abortController) {
-      this.abortController.requestCompleted = true;
-    }
+    this.abortController.requestCompleted = true;

-    /** @type {TMessage} */
    const responseMessage = {
      messageId: responseMessageId,
      conversationId,
@@ -703,32 +589,15 @@ class BaseClient {
    };

    if (typeof completion === 'string') {
-      responseMessage.text = completion;
+      responseMessage.text = addSpaceIfNeeded(generation) + completion;
    } else if (
      Array.isArray(completion) &&
-      (this.clientName === EModelEndpoint.agents ||
-        isParamEndpoint(this.options.endpoint, this.options.endpointType))
+      isParamEndpoint(this.options.endpoint, this.options.endpointType)
    ) {
      responseMessage.text = '';
-
-      if (!opts.editedContent || this.currentMessages.length === 0) {
-        responseMessage.content = completion;
-      } else {
-        const latestMessage = this.currentMessages[this.currentMessages.length - 1];
-        if (!latestMessage?.content) {
-          responseMessage.content = completion;
-        } else {
-          const existingContent = [...latestMessage.content];
-          const { type: editedType } = opts.editedContent;
-          responseMessage.content = this.mergeEditedContent(
-            existingContent,
-            completion,
-            editedType,
-          );
-        }
-      }
+      responseMessage.content = completion;
    } else if (Array.isArray(completion)) {
-      responseMessage.text = completion.join('');
+      responseMessage.text = addSpaceIfNeeded(generation) + completion.join('');
    }

    if (
@@ -749,48 +618,33 @@ class BaseClient {
      if (usage != null && Number(usage[this.outputTokensKey]) > 0) {
        responseMessage.tokenCount = usage[this.outputTokensKey];
        completionTokens = responseMessage.tokenCount;
-        await this.updateUserMessageTokenCount({
-          usage,
-          tokenCountMap,
-          userMessage,
-          userMessagePromise,
-          opts,
-        });
+        await this.updateUserMessageTokenCount({ usage, tokenCountMap, userMessage, opts });
      } else {
        responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
-        completionTokens = responseMessage.tokenCount;
-        await this.recordTokenUsage({
-          usage,
-          promptTokens,
-          completionTokens,
-          balance: balanceConfig,
-          model: responseMessage.model,
-        });
+        completionTokens = this.getTokenCount(completion);
      }
+
+      await this.recordTokenUsage({ promptTokens, completionTokens, usage });
    }

-    if (userMessagePromise) {
-      await userMessagePromise;
+    if (this.userMessagePromise) {
+      await this.userMessagePromise;
    }

    if (this.artifactPromises) {
      responseMessage.attachments = (await Promise.all(this.artifactPromises)).filter((a) => a);
    }

-    if (this.options.attachments) {
-      try {
-        saveOptions.files = this.options.attachments.map((attachments) => attachments.file_id);
-      } catch (error) {
-        logger.error('[BaseClient] Error mapping attachments for conversation', error);
-      }
-    }
-
-    responseMessage.databasePromise = this.saveMessageToDatabase(
-      responseMessage,
-      saveOptions,
-      user,
+    this.responsePromise = this.saveMessageToDatabase(responseMessage, saveOptions, user);
+    const messageCache = getLogStores(CacheKeys.MESSAGES);
+    messageCache.set(
+      responseMessageId,
+      {
+        text: responseMessage.text,
+        complete: true,
+      },
+      Time.FIVE_MINUTES,
    );
-    this.savedMessageIds.add(responseMessage.messageId);
    delete responseMessage.tokenCount;
    return responseMessage;
  }
@@ -810,16 +664,9 @@ class BaseClient {
   * @param {StreamUsage} params.usage
   * @param {Record<string, number>} params.tokenCountMap
   * @param {TMessage} params.userMessage
-   * @param {Promise<TMessage>} params.userMessagePromise
   * @param {object} params.opts
   */
-  async updateUserMessageTokenCount({
-    usage,
-    tokenCountMap,
-    userMessage,
-    userMessagePromise,
-    opts,
-  }) {
+  async updateUserMessageTokenCount({ usage, tokenCountMap, userMessage, opts }) {
    /** @type {boolean} */
    const shouldUpdateCount =
      this.calculateCurrentTokenCount != null &&
@@ -844,8 +691,7 @@ class BaseClient {

    userMessage.tokenCount = userMessageTokenCount;
    /*
-      Note: `AgentController` saves the user message if not saved here
-      (noted by `savedMessageIds`), so we update the count of its `userMessage` reference
+      Note: `AskController` saves the user message, so we update the count of its `userMessage` reference
    */
    if (typeof opts?.getReqData === 'function') {
      opts.getReqData({
@@ -854,10 +700,9 @@ class BaseClient {
    }
    /*
      Note: we update the user message to be sure it gets the calculated token count;
-      though `AgentController` saves the user message if not saved here
-      (noted by `savedMessageIds`), EditController does not
+      though `AskController` saves the user message, EditController does not
    */
-    await userMessagePromise;
+    await this.userMessagePromise;
    await this.updateMessageInDatabase({
      messageId: userMessage.messageId,
      tokenCount: userMessageTokenCount,
@@ -923,7 +768,7 @@ class BaseClient {
    }

    const savedMessage = await saveMessage(
-      this.options?.req,
+      this.options.req,
      {
        ...message,
        endpoint: this.options.endpoint,
@@ -937,40 +782,16 @@ class BaseClient {
      return { message: savedMessage };
    }

-    const fieldsToKeep = {
-      conversationId: message.conversationId,
-      endpoint: this.options.endpoint,
-      endpointType: this.options.endpointType,
-      ...endpointOptions,
-    };
-
-    const existingConvo =
-      this.fetchedConvo === true
-        ? null
-        : await getConvo(this.options?.req?.user?.id, message.conversationId);
-
-    const unsetFields = {};
-    const exceptions = new Set(['spec', 'iconURL']);
-    if (existingConvo != null) {
-      this.fetchedConvo = true;
-      for (const key in existingConvo) {
-        if (!key) {
-          continue;
-        }
-        if (excludedKeys.has(key) && !exceptions.has(key)) {
-          continue;
-        }
-
-        if (endpointOptions?.[key] === undefined) {
-          unsetFields[key] = 1;
-        }
-      }
-    }
-
-    const conversation = await saveConvo(this.options?.req, fieldsToKeep, {
-      context: 'api/app/clients/BaseClient.js - saveMessageToDatabase #saveConvo',
-      unsetFields,
-    });
+    const conversation = await saveConvo(
+      this.options.req,
+      {
+        conversationId: message.conversationId,
+        endpoint: this.options.endpoint,
+        endpointType: this.options.endpointType,
+        ...endpointOptions,
+      },
+      { context: 'api/app/clients/BaseClient.js - saveMessageToDatabase #saveConvo' },
+    );

    return { message: savedMessage, conversation };
  }
@@ -1081,9 +902,8 @@ class BaseClient {
    // Note: gpt-3.5-turbo and gpt-4 may update over time. Use default for these as well as for unknown models
    let tokensPerMessage = 3;
    let tokensPerName = 1;
-    const model = this.modelOptions?.model ?? this.model;

-    if (model === 'gpt-3.5-turbo-0301') {
+    if (this.modelOptions.model === 'gpt-3.5-turbo-0301') {
      tokensPerMessage = 4;
      tokensPerName = -1;
    }
@@ -1091,31 +911,7 @@ class BaseClient {
    const processValue = (value) => {
      if (Array.isArray(value)) {
        for (let item of value) {
-          if (
-            !item ||
-            !item.type ||
-            item.type === ContentTypes.THINK ||
-            item.type === ContentTypes.ERROR ||
-            item.type === ContentTypes.IMAGE_URL
-          ) {
-            continue;
-          }
-
-          if (item.type === ContentTypes.TOOL_CALL && item.tool_call != null) {
-            const toolName = item.tool_call?.name || '';
-            if (toolName != null && toolName && typeof toolName === 'string') {
-              numTokens += this.getTokenCount(toolName);
-            }
-
-            const args = item.tool_call?.args || '';
-            if (args != null && args && typeof args === 'string') {
-              numTokens += this.getTokenCount(args);
-            }
-
-            const output = item.tool_call?.output || '';
-            if (output != null && output && typeof output === 'string') {
-              numTokens += this.getTokenCount(output);
-            }
+          if (!item || !item.type || item.type === 'image_url') {
            continue;
          }

@@ -1147,50 +943,6 @@ class BaseClient {
    return numTokens;
  }

-  /**
-   * Merges completion content with existing content when editing TEXT or THINK types
-   * @param {Array} existingContent - The existing content array
-   * @param {Array} newCompletion - The new completion content
-   * @param {string} editedType - The type of content being edited
-   * @returns {Array} The merged content array
-   */
-  mergeEditedContent(existingContent, newCompletion, editedType) {
-    if (!newCompletion.length) {
-      return existingContent.concat(newCompletion);
-    }
-
-    if (editedType !== ContentTypes.TEXT && editedType !== ContentTypes.THINK) {
-      return existingContent.concat(newCompletion);
-    }
-
-    const lastIndex = existingContent.length - 1;
-    const lastExisting = existingContent[lastIndex];
-    const firstNew = newCompletion[0];
-
-    if (lastExisting?.type !== firstNew?.type || firstNew?.type !== editedType) {
-      return existingContent.concat(newCompletion);
-    }
-
-    const mergedContent = [...existingContent];
-    if (editedType === ContentTypes.TEXT) {
-      mergedContent[lastIndex] = {
-        ...mergedContent[lastIndex],
-        [ContentTypes.TEXT]:
-          (mergedContent[lastIndex][ContentTypes.TEXT] || '') + (firstNew[ContentTypes.TEXT] || ''),
-      };
-    } else {
-      mergedContent[lastIndex] = {
-        ...mergedContent[lastIndex],
-        [ContentTypes.THINK]:
-          (mergedContent[lastIndex][ContentTypes.THINK] || '') +
-          (firstNew[ContentTypes.THINK] || ''),
-      };
-    }
-
-    // Add remaining completion items
-    return mergedContent.concat(newCompletion.slice(1));
-  }
-
  async sendPayload(payload, opts = {}) {
    if (opts && typeof opts === 'object') {
      this.setOptions(opts);
@@ -1209,15 +961,6 @@ class BaseClient {
      return _messages;
    }

-    const seen = new Set();
-    const attachmentsProcessed =
-      this.options.attachments && !(this.options.attachments instanceof Promise);
-    if (attachmentsProcessed) {
-      for (const attachment of this.options.attachments) {
-        seen.add(attachment.file_id);
-      }
-    }
-
    /**
     *
     * @param {TMessage} message
@@ -1228,28 +971,12 @@ class BaseClient {
        this.message_file_map = {};
      }

-      const fileIds = [];
-      for (const file of message.files) {
-        if (seen.has(file.file_id)) {
-          continue;
-        }
-        fileIds.push(file.file_id);
-        seen.add(file.file_id);
-      }
+      const fileIds = message.files.map((file) => file.file_id);
+      const files = await getFiles({
+        file_id: { $in: fileIds },
+      });

-      if (fileIds.length === 0) {
-        return message;
-      }
-
-      const files = await getFiles(
-        {
-          file_id: { $in: fileIds },
-        },
-        {},
-        {},
-      );
-
-      await this.addImageURLs(message, files, this.visionMode);
+      await this.addImageURLs(message, files);

      this.message_file_map[message.messageId] = files;
      return message;
--- a/api/app/clients/ChatGPTClient.js
+++ b/api/app/clients/ChatGPTClient.js
@@ -0,0 +1,803 @@
+const Keyv = require('keyv');
+const crypto = require('crypto');
+const { CohereClient } = require('cohere-ai');
+const { fetchEventSource } = require('@waylaidwanderer/fetch-event-source');
+const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
+const {
+  ImageDetail,
+  EModelEndpoint,
+  resolveHeaders,
+  CohereConstants,
+  mapModelToAzureConfig,
+} = require('librechat-data-provider');
+const { extractBaseURL, constructAzureURL, genAzureChatCompletion } = require('~/utils');
+const { createContextHandlers } = require('./prompts');
+const { createCoherePayload } = require('./llm');
+const { Agent, ProxyAgent } = require('undici');
+const BaseClient = require('./BaseClient');
+const { logger } = require('~/config');
+
+const CHATGPT_MODEL = 'gpt-3.5-turbo';
+const tokenizersCache = {};
+
+class ChatGPTClient extends BaseClient {
+  constructor(apiKey, options = {}, cacheOptions = {}) {
+    super(apiKey, options, cacheOptions);
+
+    cacheOptions.namespace = cacheOptions.namespace || 'chatgpt';
+    this.conversationsCache = new Keyv(cacheOptions);
+    this.setOptions(options);
+  }
+
+  setOptions(options) {
+    if (this.options && !this.options.replaceOptions) {
+      // nested options aren't spread properly, so we need to do this manually
+      this.options.modelOptions = {
+        ...this.options.modelOptions,
+        ...options.modelOptions,
+      };
+      delete options.modelOptions;
+      // now we can merge options
+      this.options = {
+        ...this.options,
+        ...options,
+      };
+    } else {
+      this.options = options;
+    }
+
+    if (this.options.openaiApiKey) {
+      this.apiKey = this.options.openaiApiKey;
+    }
+
+    const modelOptions = this.options.modelOptions || {};
+    this.modelOptions = {
+      ...modelOptions,
+      // set some good defaults (check for undefined in some cases because they may be 0)
+      model: modelOptions.model || CHATGPT_MODEL,
+      temperature: typeof modelOptions.temperature === 'undefined' ? 0.8 : modelOptions.temperature,
+      top_p: typeof modelOptions.top_p === 'undefined' ? 1 : modelOptions.top_p,
+      presence_penalty:
+        typeof modelOptions.presence_penalty === 'undefined' ? 1 : modelOptions.presence_penalty,
+      stop: modelOptions.stop,
+    };
+
+    this.isChatGptModel = this.modelOptions.model.includes('gpt-');
+    const { isChatGptModel } = this;
+    this.isUnofficialChatGptModel =
+      this.modelOptions.model.startsWith('text-chat') ||
+      this.modelOptions.model.startsWith('text-davinci-002-render');
+    const { isUnofficialChatGptModel } = this;
+
+    // Davinci models have a max context length of 4097 tokens.
+    this.maxContextTokens = this.options.maxContextTokens || (isChatGptModel ? 4095 : 4097);
+    // I decided to reserve 1024 tokens for the response.
+    // The max prompt tokens is determined by the max context tokens minus the max response tokens.
+    // Earlier messages will be dropped until the prompt is within the limit.
+    this.maxResponseTokens = this.modelOptions.max_tokens || 1024;
+    this.maxPromptTokens =
+      this.options.maxPromptTokens || this.maxContextTokens - this.maxResponseTokens;
+
+    if (this.maxPromptTokens + this.maxResponseTokens > this.maxContextTokens) {
+      throw new Error(
+        `maxPromptTokens + max_tokens (${this.maxPromptTokens} + ${this.maxResponseTokens} = ${
+          this.maxPromptTokens + this.maxResponseTokens
+        }) must be less than or equal to maxContextTokens (${this.maxContextTokens})`,
+      );
+    }
+
+    this.userLabel = this.options.userLabel || 'User';
+    this.chatGptLabel = this.options.chatGptLabel || 'ChatGPT';
+
+    if (isChatGptModel) {
+      // Use these faux tokens to help the AI understand the context since we are building the chat log ourselves.
+      // Trying to use "<|im_start|>" causes the AI to still generate "<" or "<|" at the end sometimes for some reason,
+      // without tripping the stop sequences, so I'm using "||>" instead.
+      this.startToken = '||>';
+      this.endToken = '';
+      this.gptEncoder = this.constructor.getTokenizer('cl100k_base');
+    } else if (isUnofficialChatGptModel) {
+      this.startToken = '<|im_start|>';
+      this.endToken = '<|im_end|>';
+      this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true, {
+        '<|im_start|>': 100264,
+        '<|im_end|>': 100265,
+      });
+    } else {
+      // Previously I was trying to use "<|endoftext|>" but there seems to be some bug with OpenAI's token counting
+      // system that causes only the first "<|endoftext|>" to be counted as 1 token, and the rest are not treated
+      // as a single token. So we're using this instead.
+      this.startToken = '||>';
+      this.endToken = '';
+      try {
+        this.gptEncoder = this.constructor.getTokenizer(this.modelOptions.model, true);
+      } catch {
+        this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true);
+      }
+    }
+
+    if (!this.modelOptions.stop) {
+      const stopTokens = [this.startToken];
+      if (this.endToken && this.endToken !== this.startToken) {
+        stopTokens.push(this.endToken);
+      }
+      stopTokens.push(`\n${this.userLabel}:`);
+      stopTokens.push('<|diff_marker|>');
+      // I chose not to do one for `chatGptLabel` because I've never seen it happen
+      this.modelOptions.stop = stopTokens;
+    }
+
+    if (this.options.reverseProxyUrl) {
+      this.completionsUrl = this.options.reverseProxyUrl;
+    } else if (isChatGptModel) {
+      this.completionsUrl = 'https://api.openai.com/v1/chat/completions';
+    } else {
+      this.completionsUrl = 'https://api.openai.com/v1/completions';
+    }
+
+    return this;
+  }
+
+  static getTokenizer(encoding, isModelName = false, extendSpecialTokens = {}) {
+    if (tokenizersCache[encoding]) {
+      return tokenizersCache[encoding];
+    }
+    let tokenizer;
+    if (isModelName) {
+      tokenizer = encodingForModel(encoding, extendSpecialTokens);
+    } else {
+      tokenizer = getEncoding(encoding, extendSpecialTokens);
+    }
+    tokenizersCache[encoding] = tokenizer;
+    return tokenizer;
+  }
+
+  /** @type {getCompletion} */
+  async getCompletion(input, onProgress, onTokenProgress, abortController = null) {
+    if (!abortController) {
+      abortController = new AbortController();
+    }
+
+    let modelOptions = { ...this.modelOptions };
+    if (typeof onProgress === 'function') {
+      modelOptions.stream = true;
+    }
+    if (this.isChatGptModel) {
+      modelOptions.messages = input;
+    } else {
+      modelOptions.prompt = input;
+    }
+
+    if (this.useOpenRouter && modelOptions.prompt) {
+      delete modelOptions.stop;
+    }
+
+    const { debug } = this.options;
+    let baseURL = this.completionsUrl;
+    if (debug) {
+      console.debug();
+      console.debug(baseURL);
+      console.debug(modelOptions);
+      console.debug();
+    }
+
+    const opts = {
+      method: 'POST',
+      headers: {
+        'Content-Type': 'application/json',
+      },
+      dispatcher: new Agent({
+        bodyTimeout: 0,
+        headersTimeout: 0,
+      }),
+    };
+
+    if (this.isVisionModel) {
+      modelOptions.max_tokens = 4000;
+    }
+
+    /** @type {TAzureConfig | undefined} */
+    const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
+
+    const isAzure = this.azure || this.options.azure;
+    if (
+      (isAzure && this.isVisionModel && azureConfig) ||
+      (azureConfig && this.isVisionModel && this.options.endpoint === EModelEndpoint.azureOpenAI)
+    ) {
+      const { modelGroupMap, groupMap } = azureConfig;
+      const {
+        azureOptions,
+        baseURL,
+        headers = {},
+        serverless,
+      } = mapModelToAzureConfig({
+        modelName: modelOptions.model,
+        modelGroupMap,
+        groupMap,
+      });
+      opts.headers = resolveHeaders(headers);
+      this.langchainProxy = extractBaseURL(baseURL);
+      this.apiKey = azureOptions.azureOpenAIApiKey;
+
+      const groupName = modelGroupMap[modelOptions.model].group;
+      this.options.addParams = azureConfig.groupMap[groupName].addParams;
+      this.options.dropParams = azureConfig.groupMap[groupName].dropParams;
+      // Note: `forcePrompt` not re-assigned as only chat models are vision models
+
+      this.azure = !serverless && azureOptions;
+      this.azureEndpoint =
+        !serverless && genAzureChatCompletion(this.azure, modelOptions.model, this);
+    }
+
+    if (this.options.headers) {
+      opts.headers = { ...opts.headers, ...this.options.headers };
+    }
+
+    if (isAzure) {
+      // Azure does not accept `model` in the body, so we need to remove it.
+      delete modelOptions.model;
+
+      baseURL = this.langchainProxy
+        ? constructAzureURL({
+          baseURL: this.langchainProxy,
+          azureOptions: this.azure,
+        })
+        : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];
+
+      if (this.options.forcePrompt) {
+        baseURL += '/completions';
+      } else {
+        baseURL += '/chat/completions';
+      }
+
+      opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
+      opts.headers = { ...opts.headers, 'api-key': this.apiKey };
+    } else if (this.apiKey) {
+      opts.headers.Authorization = `Bearer ${this.apiKey}`;
+    }
+
+    if (process.env.OPENAI_ORGANIZATION) {
+      opts.headers['OpenAI-Organization'] = process.env.OPENAI_ORGANIZATION;
+    }
+
+    if (this.useOpenRouter) {
+      opts.headers['HTTP-Referer'] = 'https://librechat.ai';
+      opts.headers['X-Title'] = 'LibreChat';
+    }
+
+    if (this.options.proxy) {
+      opts.dispatcher = new ProxyAgent(this.options.proxy);
+    }
+
+    /* hacky fixes for Mistral AI API:
+      - Re-orders system message to the top of the messages payload, as not allowed anywhere else
+      - If there is only one message and it's a system message, change the role to user
+      */
+    if (baseURL.includes('https://api.mistral.ai/v1') && modelOptions.messages) {
+      const { messages } = modelOptions;
+
+      const systemMessageIndex = messages.findIndex((msg) => msg.role === 'system');
+
+      if (systemMessageIndex > 0) {
+        const [systemMessage] = messages.splice(systemMessageIndex, 1);
+        messages.unshift(systemMessage);
+      }
+
+      modelOptions.messages = messages;
+
+      if (messages.length === 1 && messages[0].role === 'system') {
+        modelOptions.messages[0].role = 'user';
+      }
+    }
+
+    if (this.options.addParams && typeof this.options.addParams === 'object') {
+      modelOptions = {
+        ...modelOptions,
+        ...this.options.addParams,
+      };
+      logger.debug('[ChatGPTClient] chatCompletion: added params', {
+        addParams: this.options.addParams,
+        modelOptions,
+      });
+    }
+
+    if (this.options.dropParams && Array.isArray(this.options.dropParams)) {
+      this.options.dropParams.forEach((param) => {
+        delete modelOptions[param];
+      });
+      logger.debug('[ChatGPTClient] chatCompletion: dropped params', {
+        dropParams: this.options.dropParams,
+        modelOptions,
+      });
+    }
+
+    if (baseURL.startsWith(CohereConstants.API_URL)) {
+      const payload = createCoherePayload({ modelOptions });
+      return await this.cohereChatCompletion({ payload, onTokenProgress });
+    }
+
+    if (baseURL.includes('v1') && !baseURL.includes('/completions') && !this.isChatCompletion) {
+      baseURL = baseURL.split('v1')[0] + 'v1/completions';
+    } else if (
+      baseURL.includes('v1') &&
+      !baseURL.includes('/chat/completions') &&
+      this.isChatCompletion
+    ) {
+      baseURL = baseURL.split('v1')[0] + 'v1/chat/completions';
+    }
+
+    const BASE_URL = new URL(baseURL);
+    if (opts.defaultQuery) {
+      Object.entries(opts.defaultQuery).forEach(([key, value]) => {
+        BASE_URL.searchParams.append(key, value);
+      });
+      delete opts.defaultQuery;
+    }
+
+    const completionsURL = BASE_URL.toString();
+    opts.body = JSON.stringify(modelOptions);
+
+    if (modelOptions.stream) {
+      // eslint-disable-next-line no-async-promise-executor
+      return new Promise(async (resolve, reject) => {
+        try {
+          let done = false;
+          await fetchEventSource(completionsURL, {
+            ...opts,
+            signal: abortController.signal,
+            async onopen(response) {
+              if (response.status === 200) {
+                return;
+              }
+              if (debug) {
+                console.debug(response);
+              }
+              let error;
+              try {
+                const body = await response.text();
+                error = new Error(`Failed to send message. HTTP ${response.status} - ${body}`);
+                error.status = response.status;
+                error.json = JSON.parse(body);
+              } catch {
+                error = error || new Error(`Failed to send message. HTTP ${response.status}`);
+              }
+              throw error;
+            },
+            onclose() {
+              if (debug) {
+                console.debug('Server closed the connection unexpectedly, returning...');
+              }
+              // workaround for private API not sending [DONE] event
+              if (!done) {
+                onProgress('[DONE]');
+                resolve();
+              }
+            },
+            onerror(err) {
+              if (debug) {
+                console.debug(err);
+              }
+              // rethrow to stop the operation
+              throw err;
+            },
+            onmessage(message) {
+              if (debug) {
+                console.debug(message);
+              }
+              if (!message.data || message.event === 'ping') {
+                return;
+              }
+              if (message.data === '[DONE]') {
+                onProgress('[DONE]');
+                resolve();
+                done = true;
+                return;
+              }
+              onProgress(JSON.parse(message.data));
+            },
+          });
+        } catch (err) {
+          reject(err);
+        }
+      });
+    }
+    const response = await fetch(completionsURL, {
+      ...opts,
+      signal: abortController.signal,
+    });
+    if (response.status !== 200) {
+      const body = await response.text();
+      const error = new Error(`Failed to send message. HTTP ${response.status} - ${body}`);
+      error.status = response.status;
+      try {
+        error.json = JSON.parse(body);
+      } catch {
+        error.body = body;
+      }
+      throw error;
+    }
+    return response.json();
+  }
+
+  /** @type {cohereChatCompletion} */
+  async cohereChatCompletion({ payload, onTokenProgress }) {
+    const cohere = new CohereClient({
+      token: this.apiKey,
+      environment: this.completionsUrl,
+    });
+
+    if (!payload.stream) {
+      const chatResponse = await cohere.chat(payload);
+      return chatResponse.text;
+    }
+
+    const chatStream = await cohere.chatStream(payload);
+    let reply = '';
+    for await (const message of chatStream) {
+      if (!message) {
+        continue;
+      }
+
+      if (message.eventType === 'text-generation' && message.text) {
+        onTokenProgress(message.text);
+        reply += message.text;
+      }
+      /*
+      Cohere API Chinese Unicode character replacement hotfix.
+      Should be un-commented when the following issue is resolved:
+      https://github.com/cohere-ai/cohere-typescript/issues/151
+
+      else if (message.eventType === 'stream-end' && message.response) {
+        reply = message.response.text;
+      }
+      */
+    }
+
+    return reply;
+  }
+
+  async generateTitle(userMessage, botMessage) {
+    const instructionsPayload = {
+      role: 'system',
+      content: `Write an extremely concise subtitle for this conversation with no more than a few words. All words should be capitalized. Exclude punctuation.
+
+||>Message:
+${userMessage.message}
+||>Response:
+${botMessage.message}
+
+||>Title:`,
+    };
+
+    const titleGenClientOptions = JSON.parse(JSON.stringify(this.options));
+    titleGenClientOptions.modelOptions = {
+      model: 'gpt-3.5-turbo',
+      temperature: 0,
+      presence_penalty: 0,
+      frequency_penalty: 0,
+    };
+    const titleGenClient = new ChatGPTClient(this.apiKey, titleGenClientOptions);
+    const result = await titleGenClient.getCompletion([instructionsPayload], null);
+    // remove any non-alphanumeric characters, replace multiple spaces with 1, and then trim
+    return result.choices[0].message.content
+      .replace(/[^a-zA-Z0-9' ]/g, '')
+      .replace(/\s+/g, ' ')
+      .trim();
+  }
+
+  async sendMessage(message, opts = {}) {
+    if (opts.clientOptions && typeof opts.clientOptions === 'object') {
+      this.setOptions(opts.clientOptions);
+    }
+
+    const conversationId = opts.conversationId || crypto.randomUUID();
+    const parentMessageId = opts.parentMessageId || crypto.randomUUID();
+
+    let conversation =
+      typeof opts.conversation === 'object'
+        ? opts.conversation
+        : await this.conversationsCache.get(conversationId);
+
+    let isNewConversation = false;
+    if (!conversation) {
+      conversation = {
+        messages: [],
+        createdAt: Date.now(),
+      };
+      isNewConversation = true;
+    }
+
+    const shouldGenerateTitle = opts.shouldGenerateTitle && isNewConversation;
+
+    const userMessage = {
+      id: crypto.randomUUID(),
+      parentMessageId,
+      role: 'User',
+      message,
+    };
+    conversation.messages.push(userMessage);
+
+    // Doing it this way instead of having each message be a separate element in the array seems to be more reliable,
+    // especially when it comes to keeping the AI in character. It also seems to improve coherency and context retention.
+    const { prompt: payload, context } = await this.buildPrompt(
+      conversation.messages,
+      userMessage.id,
+      {
+        isChatGptModel: this.isChatGptModel,
+        promptPrefix: opts.promptPrefix,
+      },
+    );
+
+    if (this.options.keepNecessaryMessagesOnly) {
+      conversation.messages = context;
+    }
+
+    let reply = '';
+    let result = null;
+    if (typeof opts.onProgress === 'function') {
+      await this.getCompletion(
+        payload,
+        (progressMessage) => {
+          if (progressMessage === '[DONE]') {
+            return;
+          }
+          const token = this.isChatGptModel
+            ? progressMessage.choices[0].delta.content
+            : progressMessage.choices[0].text;
+          // first event's delta content is always undefined
+          if (!token) {
+            return;
+          }
+          if (this.options.debug) {
+            console.debug(token);
+          }
+          if (token === this.endToken) {
+            return;
+          }
+          opts.onProgress(token);
+          reply += token;
+        },
+        opts.abortController || new AbortController(),
+      );
+    } else {
+      result = await this.getCompletion(
+        payload,
+        null,
+        opts.abortController || new AbortController(),
+      );
+      if (this.options.debug) {
+        console.debug(JSON.stringify(result));
+      }
+      if (this.isChatGptModel) {
+        reply = result.choices[0].message.content;
+      } else {
+        reply = result.choices[0].text.replace(this.endToken, '');
+      }
+    }
+
+    // avoids some rendering issues when using the CLI app
+    if (this.options.debug) {
+      console.debug();
+    }
+
+    reply = reply.trim();
+
+    const replyMessage = {
+      id: crypto.randomUUID(),
+      parentMessageId: userMessage.id,
+      role: 'ChatGPT',
+      message: reply,
+    };
+    conversation.messages.push(replyMessage);
+
+    const returnData = {
+      response: replyMessage.message,
+      conversationId,
+      parentMessageId: replyMessage.parentMessageId,
+      messageId: replyMessage.id,
+      details: result || {},
+    };
+
+    if (shouldGenerateTitle) {
+      conversation.title = await this.generateTitle(userMessage, replyMessage);
+      returnData.title = conversation.title;
+    }
+
+    await this.conversationsCache.set(conversationId, conversation);
+
+    if (this.options.returnConversation) {
+      returnData.conversation = conversation;
+    }
+
+    return returnData;
+  }
+
+  async buildPrompt(messages, { isChatGptModel = false, promptPrefix = null }) {
+    promptPrefix = (promptPrefix || this.options.promptPrefix || '').trim();
+
+    // Handle attachments and create augmentedPrompt
+    if (this.options.attachments) {
+      const attachments = await this.options.attachments;
+      const lastMessage = messages[messages.length - 1];
+
+      if (this.message_file_map) {
+        this.message_file_map[lastMessage.messageId] = attachments;
+      } else {
+        this.message_file_map = {
+          [lastMessage.messageId]: attachments,
+        };
+      }
+
+      const files = await this.addImageURLs(lastMessage, attachments);
+      this.options.attachments = files;
+
+      this.contextHandlers = createContextHandlers(this.options.req, lastMessage.text);
+    }
+
+    if (this.message_file_map) {
+      this.contextHandlers = createContextHandlers(
+        this.options.req,
+        messages[messages.length - 1].text,
+      );
+    }
+
+    // Calculate image token cost and process embedded files
+    messages.forEach((message, i) => {
+      if (this.message_file_map && this.message_file_map[message.messageId]) {
+        const attachments = this.message_file_map[message.messageId];
+        for (const file of attachments) {
+          if (file.embedded) {
+            this.contextHandlers?.processFile(file);
+            continue;
+          }
+
+          messages[i].tokenCount =
+            (messages[i].tokenCount || 0) +
+            this.calculateImageTokenCost({
+              width: file.width,
+              height: file.height,
+              detail: this.options.imageDetail ?? ImageDetail.auto,
+            });
+        }
+      }
+    });
+
+    if (this.contextHandlers) {
+      this.augmentedPrompt = await this.contextHandlers.createContext();
+      promptPrefix = this.augmentedPrompt + promptPrefix;
+    }
+
+    if (promptPrefix) {
+      // If the prompt prefix doesn't end with the end token, add it.
+      if (!promptPrefix.endsWith(`${this.endToken}`)) {
+        promptPrefix = `${promptPrefix.trim()}${this.endToken}\n\n`;
+      }
+      promptPrefix = `${this.startToken}Instructions:\n${promptPrefix}`;
+    }
+    const promptSuffix = `${this.startToken}${this.chatGptLabel}:\n`; // Prompt ChatGPT to respond.
+
+    const instructionsPayload = {
+      role: 'system',
+      content: promptPrefix,
+    };
+
+    const messagePayload = {
+      role: 'system',
+      content: promptSuffix,
+    };
+
+    let currentTokenCount;
+    if (isChatGptModel) {
+      currentTokenCount =
+        this.getTokenCountForMessage(instructionsPayload) +
+        this.getTokenCountForMessage(messagePayload);
+    } else {
+      currentTokenCount = this.getTokenCount(`${promptPrefix}${promptSuffix}`);
+    }
+    let promptBody = '';
+    const maxTokenCount = this.maxPromptTokens;
+
+    const context = [];
+
+    // Iterate backwards through the messages, adding them to the prompt until we reach the max token count.
+    // Do this within a recursive async function so that it doesn't block the event loop for too long.
+    const buildPromptBody = async () => {
+      if (currentTokenCount < maxTokenCount && messages.length > 0) {
+        const message = messages.pop();
+        const roleLabel =
+          message?.isCreatedByUser || message?.role?.toLowerCase() === 'user'
+            ? this.userLabel
+            : this.chatGptLabel;
+        const messageString = `${this.startToken}${roleLabel}:\n${
+          message?.text ?? message?.message
+        }${this.endToken}\n`;
+        let newPromptBody;
+        if (promptBody || isChatGptModel) {
+          newPromptBody = `${messageString}${promptBody}`;
+        } else {
+          // Always insert prompt prefix before the last user message, if not gpt-3.5-turbo.
+          // This makes the AI obey the prompt instructions better, which is important for custom instructions.
+          // After a bunch of testing, it doesn't seem to cause the AI any confusion, even if you ask it things
+          // like "what's the last thing I wrote?".
+          newPromptBody = `${promptPrefix}${messageString}${promptBody}`;
+        }
+
+        context.unshift(message);
+
+        const tokenCountForMessage = this.getTokenCount(messageString);
+        const newTokenCount = currentTokenCount + tokenCountForMessage;
+        if (newTokenCount > maxTokenCount) {
+          if (promptBody) {
+            // This message would put us over the token limit, so don't add it.
+            return false;
+          }
+          // This is the first message, so we can't add it. Just throw an error.
+          throw new Error(
+            `Prompt is too long. Max token count is ${maxTokenCount}, but prompt is ${newTokenCount} tokens long.`,
+          );
+        }
+        promptBody = newPromptBody;
+        currentTokenCount = newTokenCount;
+        // wait for next tick to avoid blocking the event loop
+        await new Promise((resolve) => setImmediate(resolve));
+        return buildPromptBody();
+      }
+      return true;
+    };
+
+    await buildPromptBody();
+
+    const prompt = `${promptBody}${promptSuffix}`;
+    if (isChatGptModel) {
+      messagePayload.content = prompt;
+      // Add 3 tokens for Assistant Label priming after all messages have been counted.
+      currentTokenCount += 3;
+    }
+
+    // Use up to `this.maxContextTokens` tokens (prompt + response), but try to leave `this.maxTokens` tokens for the response.
+    this.modelOptions.max_tokens = Math.min(
+      this.maxContextTokens - currentTokenCount,
+      this.maxResponseTokens,
+    );
+
+    if (isChatGptModel) {
+      return { prompt: [instructionsPayload, messagePayload], context };
+    }
+    return { prompt, context, promptTokens: currentTokenCount };
+  }
+
+  getTokenCount(text) {
+    return this.gptEncoder.encode(text, 'all').length;
+  }
+
+  /**
+   * Algorithm adapted from "6. Counting tokens for chat API calls" of
+   * https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb
+   *
+   * An additional 3 tokens need to be added for assistant label priming after all messages have been counted.
+   *
+   * @param {Object} message
+   */
+  getTokenCountForMessage(message) {
+    // Note: gpt-3.5-turbo and gpt-4 may update over time. Use default for these as well as for unknown models
+    let tokensPerMessage = 3;
+    let tokensPerName = 1;
+
+    if (this.modelOptions.model === 'gpt-3.5-turbo-0301') {
+      tokensPerMessage = 4;
+      tokensPerName = -1;
+    }
+
+    let numTokens = tokensPerMessage;
+    for (let [key, value] of Object.entries(message)) {
+      numTokens += this.getTokenCount(value);
+      if (key === 'name') {
+        numTokens += tokensPerName;
+      }
+    }
+
+    return numTokens;
+  }
+}
+
+module.exports = ChatGPTClient;
--- a/api/app/clients/GoogleClient.js
+++ b/api/app/clients/GoogleClient.js
@@ -1,27 +1,23 @@
 const { google } = require('googleapis');
-const { getModelMaxTokens } = require('@librechat/api');
-const { concat } = require('@langchain/core/utils/stream');
+const { Agent, ProxyAgent } = require('undici');
 const { ChatVertexAI } = require('@langchain/google-vertexai');
-const { Tokenizer, getSafetySettings } = require('@librechat/api');
+const { GoogleVertexAI } = require('@langchain/google-vertexai');
+const { ChatGoogleVertexAI } = require('@langchain/google-vertexai');
 const { ChatGoogleGenerativeAI } = require('@langchain/google-genai');
 const { GoogleGenerativeAI: GenAI } = require('@google/generative-ai');
-const { HumanMessage, SystemMessage } = require('@langchain/core/messages');
+const { AIMessage, HumanMessage, SystemMessage } = require('@langchain/core/messages');
+const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
 const {
-  googleGenConfigSchema,
  validateVisionModel,
  getResponseSender,
  endpointSettings,
-  parseTextParts,
  EModelEndpoint,
-  googleSettings,
-  ContentTypes,
  VisionModes,
-  ErrorTypes,
  Constants,
  AuthKeys,
 } = require('librechat-data-provider');
 const { encodeAndFormat } = require('~/server/services/Files/images');
-const { spendTokens } = require('~/models/spendTokens');
+const { getModelMaxTokens } = require('~/utils');
 const { sleep } = require('~/server/utils');
 const { logger } = require('~/config');
 const {
@@ -34,11 +30,11 @@ const BaseClient = require('./BaseClient');

 const loc = process.env.GOOGLE_LOC || 'us-central1';
 const publisher = 'google';
-const endpointPrefix =
-  loc === 'global' ? 'aiplatform.googleapis.com' : `${loc}-aiplatform.googleapis.com`;
+const endpointPrefix = `https://${loc}-aiplatform.googleapis.com`;
+// const apiEndpoint = loc + '-aiplatform.googleapis.com';
+const tokenizersCache = {};

 const settings = endpointSettings[EModelEndpoint.google];
-const EXCLUDED_GENAI_MODELS = /gemini-(?:1\.0|1-0|pro)/;

 class GoogleClient extends BaseClient {
  constructor(credentials, options = {}) {
@@ -53,30 +49,14 @@ class GoogleClient extends BaseClient {

    const serviceKey = creds[AuthKeys.GOOGLE_SERVICE_KEY] ?? {};
    this.serviceKey =
-      serviceKey && typeof serviceKey === 'string' ? JSON.parse(serviceKey) : (serviceKey ?? {});
-    /** @type {string | null | undefined} */
-    this.project_id = this.serviceKey.project_id;
+      serviceKey && typeof serviceKey === 'string' ? JSON.parse(serviceKey) : serviceKey ?? {};
    this.client_email = this.serviceKey.client_email;
    this.private_key = this.serviceKey.private_key;
+    this.project_id = this.serviceKey.project_id;
    this.access_token = null;

    this.apiKey = creds[AuthKeys.GOOGLE_API_KEY];

-    this.reverseProxyUrl = options.reverseProxyUrl;
-
-    this.authHeader = options.authHeader;
-
-    /** @type {UsageMetadata | undefined} */
-    this.usage;
-    /** The key for the usage object's input tokens
-     * @type {string} */
-    this.inputTokensKey = 'input_tokens';
-    /** The key for the usage object's output tokens
-     * @type {string} */
-    this.outputTokensKey = 'output_tokens';
-    this.visionMode = VisionModes.generative;
-    /** @type {string} */
-    this.systemMessage;
    if (options.skipSetOptions) {
      return;
    }
@@ -85,7 +65,7 @@ class GoogleClient extends BaseClient {

  /* Google specific methods */
  constructUrl() {
-    return `https://${endpointPrefix}/v1/projects/${this.project_id}/locations/${loc}/publishers/${publisher}/models/${this.modelOptions.model}:serverStreamingPredict`;
+    return `${endpointPrefix}/v1/projects/${this.project_id}/locations/${loc}/publishers/${publisher}/models/${this.modelOptions.model}:serverStreamingPredict`;
  }

  async getClient() {
@@ -136,12 +116,22 @@ class GoogleClient extends BaseClient {
      this.options = options;
    }

+    this.options.examples = (this.options.examples ?? [])
+      .filter((ex) => ex)
+      .filter((obj) => obj.input.content !== '' && obj.output.content !== '');
+
    this.modelOptions = this.options.modelOptions || {};

    this.options.attachments?.then((attachments) => this.checkVisionRequest(attachments));

    /** @type {boolean} Whether using a "GenerativeAI" Model */
-    this.isGenerativeModel = /gemini|learnlm|gemma/.test(this.modelOptions.model);
+    this.isGenerativeModel = this.modelOptions.model.includes('gemini');
+    const { isGenerativeModel } = this;
+    this.isChatModel = !isGenerativeModel && this.modelOptions.model.includes('chat');
+    const { isChatModel } = this;
+    this.isTextModel =
+      !isGenerativeModel && !isChatModel && /code|text/.test(this.modelOptions.model);
+    const { isTextModel } = this;

    this.maxContextTokens =
      this.options.maxContextTokens ??
@@ -166,16 +156,6 @@ class GoogleClient extends BaseClient {
      );
    }

-    // Add thinking configuration
-    this.modelOptions.thinkingConfig = {
-      thinkingBudget:
-        (this.modelOptions.thinking ?? googleSettings.thinking.default)
-          ? this.modelOptions.thinkingBudget
-          : 0,
-    };
-    delete this.modelOptions.thinking;
-    delete this.modelOptions.thinkingBudget;
-
    this.sender =
      this.options.sender ??
      getResponseSender({
@@ -187,18 +167,50 @@ class GoogleClient extends BaseClient {
    this.userLabel = this.options.userLabel || 'User';
    this.modelLabel = this.options.modelLabel || 'Assistant';

+    if (isChatModel || isGenerativeModel) {
+      // Use these faux tokens to help the AI understand the context since we are building the chat log ourselves.
+      // Trying to use "<|im_start|>" causes the AI to still generate "<" or "<|" at the end sometimes for some reason,
+      // without tripping the stop sequences, so I'm using "||>" instead.
+      this.startToken = '||>';
+      this.endToken = '';
+      this.gptEncoder = this.constructor.getTokenizer('cl100k_base');
+    } else if (isTextModel) {
+      this.startToken = '||>';
+      this.endToken = '';
+      this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true, {
+        '<|im_start|>': 100264,
+        '<|im_end|>': 100265,
+      });
+    } else {
+      // Previously I was trying to use "<|endoftext|>" but there seems to be some bug with OpenAI's token counting
+      // system that causes only the first "<|endoftext|>" to be counted as 1 token, and the rest are not treated
+      // as a single token. So we're using this instead.
+      this.startToken = '||>';
+      this.endToken = '';
+      try {
+        this.gptEncoder = this.constructor.getTokenizer(this.modelOptions.model, true);
+      } catch {
+        this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true);
+      }
+    }
+
+    if (!this.modelOptions.stop) {
+      const stopTokens = [this.startToken];
+      if (this.endToken && this.endToken !== this.startToken) {
+        stopTokens.push(this.endToken);
+      }
+      stopTokens.push(`\n${this.userLabel}:`);
+      stopTokens.push('<|diff_marker|>');
+      // I chose not to do one for `modelLabel` because I've never seen it happen
+      this.modelOptions.stop = stopTokens;
+    }
+
    if (this.options.reverseProxyUrl) {
      this.completionsUrl = this.options.reverseProxyUrl;
    } else {
      this.completionsUrl = this.constructUrl();
    }

-    let promptPrefix = (this.options.promptPrefix ?? '').trim();
-    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
-      promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
-    }
-    this.systemMessage = promptPrefix;
-    this.initializeClient();
    return this;
  }

@@ -209,11 +221,7 @@ class GoogleClient extends BaseClient {
   */
  checkVisionRequest(attachments) {
    /* Validation vision request */
-    this.defaultVisionModel =
-      this.options.visionModel ??
-      (!EXCLUDED_GENAI_MODELS.test(this.modelOptions.model)
-        ? this.modelOptions.model
-        : 'gemini-pro-vision');
+    this.defaultVisionModel = this.options.visionModel ?? 'gemini-pro-vision';
    const availableModels = this.options.modelsConfig?.[EModelEndpoint.google];
    this.isVisionModel = validateVisionModel({ model: this.modelOptions.model, availableModels });

@@ -234,29 +242,10 @@ class GoogleClient extends BaseClient {
  }

  formatMessages() {
-    return ((message) => {
-      const msg = {
-        author: message?.author ?? (message.isCreatedByUser ? this.userLabel : this.modelLabel),
-        content: message?.content ?? message.text,
-      };
-
-      if (!message.image_urls?.length) {
-        return msg;
-      }
-
-      msg.content = (
-        !Array.isArray(msg.content)
-          ? [
-              {
-                type: ContentTypes.TEXT,
-                [ContentTypes.TEXT]: msg.content,
-              },
-            ]
-          : msg.content
-      ).concat(message.image_urls);
-
-      return msg;
-    }).bind(this);
+    return ((message) => ({
+      author: message?.author ?? (message.isCreatedByUser ? this.userLabel : this.modelLabel),
+      content: message?.content ?? message.text,
+    })).bind(this);
  }

  /**
@@ -328,13 +317,10 @@ class GoogleClient extends BaseClient {
          this.contextHandlers?.processFile(file);
          continue;
        }
-        if (file.metadata?.fileIdentifier) {
-          continue;
-        }
      }

      this.augmentedPrompt = await this.contextHandlers.createContext();
-      this.systemMessage = this.augmentedPrompt + this.systemMessage;
+      this.options.promptPrefix = this.augmentedPrompt + this.options.promptPrefix;
    }
  }

@@ -357,6 +343,7 @@ class GoogleClient extends BaseClient {
          messages: [new HumanMessage(formatMessage({ message: latestMessage }))],
        },
      ],
+      parameters: this.modelOptions,
    };
    return { prompt: payload };
  }
@@ -372,58 +359,23 @@ class GoogleClient extends BaseClient {
    return { prompt: formattedMessages };
  }

-  /**
-   * @param {TMessage[]} [messages=[]]
-   * @param {string} [parentMessageId]
-   */
-  async buildMessages(_messages = [], parentMessageId) {
+  async buildMessages(messages = [], parentMessageId) {
    if (!this.isGenerativeModel && !this.project_id) {
-      throw new Error('[GoogleClient] PaLM 2 and Codey models are no longer supported.');
+      throw new Error(
+        '[GoogleClient] a Service Account JSON Key is required for PaLM 2 and Codey models (Vertex AI)',
+      );
    }

-    if (this.systemMessage) {
-      const instructionsTokenCount = this.getTokenCount(this.systemMessage);
-
-      this.maxContextTokens = this.maxContextTokens - instructionsTokenCount;
-      if (this.maxContextTokens < 0) {
-        const info = `${instructionsTokenCount} / ${this.maxContextTokens}`;
-        const errorMessage = `{ "type": "${ErrorTypes.INPUT_LENGTH}", "info": "${info}" }`;
-        logger.warn(`Instructions token count exceeds max context (${info}).`);
-        throw new Error(errorMessage);
-      }
-    }
-
-    for (let i = 0; i < _messages.length; i++) {
-      const message = _messages[i];
-      if (!message.tokenCount) {
-        _messages[i].tokenCount = this.getTokenCountForMessage({
-          role: message.isCreatedByUser ? 'user' : 'assistant',
-          content: message.content ?? message.text,
-        });
-      }
-    }
-
-    const {
-      payload: messages,
-      tokenCountMap,
-      promptTokens,
-    } = await this.handleContextStrategy({
-      orderedMessages: _messages,
-      formattedMessages: _messages,
-    });
-
-    if (!this.project_id && !EXCLUDED_GENAI_MODELS.test(this.modelOptions.model)) {
-      const result = await this.buildGenerativeMessages(messages);
-      result.tokenCountMap = tokenCountMap;
-      result.promptTokens = promptTokens;
-      return result;
+    if (!this.project_id && this.modelOptions.model.includes('1.5')) {
+      return await this.buildGenerativeMessages(messages);
    }

    if (this.options.attachments && this.isGenerativeModel) {
-      const result = this.buildVisionMessages(messages, parentMessageId);
-      result.tokenCountMap = tokenCountMap;
-      result.promptTokens = promptTokens;
-      return result;
+      return this.buildVisionMessages(messages, parentMessageId);
+    }
+
+    if (this.isTextModel) {
+      return this.buildMessagesPrompt(messages, parentMessageId);
    }

    let payload = {
@@ -435,14 +387,25 @@ class GoogleClient extends BaseClient {
            .map((message) => formatMessage({ message, langChain: true })),
        },
      ],
+      parameters: this.modelOptions,
    };

-    if (this.systemMessage) {
-      payload.instances[0].context = this.systemMessage;
+    let promptPrefix = (this.options.promptPrefix ?? '').trim();
+    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+      promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
+    }
+
+    if (promptPrefix) {
+      payload.instances[0].context = promptPrefix;
+    }
+
+    if (this.options.examples.length > 0) {
+      payload.instances[0].examples = this.options.examples;
    }

    logger.debug('[GoogleClient] buildMessages', payload);
-    return { prompt: payload, tokenCountMap, promptTokens };
+
+    return { prompt: payload };
  }

  async buildMessagesPrompt(messages, parentMessageId) {
@@ -456,7 +419,10 @@ class GoogleClient extends BaseClient {
      parentMessageId,
    });

-    const formattedMessages = orderedMessages.map(this.formatMessages());
+    const formattedMessages = orderedMessages.map((message) => ({
+      author: message.isCreatedByUser ? this.userLabel : this.modelLabel,
+      content: message?.content ?? message.text,
+    }));

    let lastAuthor = '';
    let groupedMessages = [];
@@ -484,7 +450,17 @@ class GoogleClient extends BaseClient {
      identityPrefix = `${identityPrefix}\nYou are ${this.options.modelLabel}`;
    }

-    let promptPrefix = (this.systemMessage ?? '').trim();
+    let promptPrefix = (this.options.promptPrefix ?? '').trim();
+    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+      promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
+    }
+    if (promptPrefix) {
+      // If the prompt prefix doesn't end with the end token, add it.
+      if (!promptPrefix.endsWith(`${this.endToken}`)) {
+        promptPrefix = `${promptPrefix.trim()}${this.endToken}\n\n`;
+      }
+      promptPrefix = `\nContext:\n${promptPrefix}`;
+    }

    if (identityPrefix) {
      promptPrefix = `${identityPrefix}${promptPrefix}`;
@@ -521,7 +497,7 @@ class GoogleClient extends BaseClient {
          isCreatedByUser || !isEdited
            ? `\n\n${message.author}:`
            : `${promptPrefix}\n\n${message.author}:`;
-        const messageString = `${messagePrefix}\n${message.content}\n`;
+        const messageString = `${messagePrefix}\n${message.content}${this.endToken}\n`;
        let newPromptBody = `${messageString}${promptBody}`;

        context.unshift(message);
@@ -587,48 +563,71 @@ class GoogleClient extends BaseClient {
    return { prompt, context };
  }

+  async _getCompletion(payload, abortController = null) {
+    if (!abortController) {
+      abortController = new AbortController();
+    }
+    const { debug } = this.options;
+    const url = this.completionsUrl;
+    if (debug) {
+      logger.debug('GoogleClient _getCompletion', { url, payload });
+    }
+    const opts = {
+      method: 'POST',
+      agent: new Agent({
+        bodyTimeout: 0,
+        headersTimeout: 0,
+      }),
+      signal: abortController.signal,
+    };
+
+    if (this.options.proxy) {
+      opts.agent = new ProxyAgent(this.options.proxy);
+    }
+
+    const client = await this.getClient();
+    const res = await client.request({ url, method: 'POST', data: payload });
+    logger.debug('GoogleClient _getCompletion', { res });
+    return res.data;
+  }
+
  createLLM(clientOptions) {
    const model = clientOptions.modelName ?? clientOptions.model;
    clientOptions.location = loc;
-    clientOptions.endpoint = endpointPrefix;
-
-    let requestOptions = null;
-    if (this.reverseProxyUrl) {
-      requestOptions = {
-        baseUrl: this.reverseProxyUrl,
-      };
-
-      if (this.authHeader) {
-        requestOptions.customHeaders = {
-          Authorization: `Bearer ${this.apiKey}`,
-        };
-      }
-    }
-
-    if (this.project_id != null) {
+    clientOptions.endpoint = `${loc}-aiplatform.googleapis.com`;
+    if (this.project_id && this.isTextModel) {
+      logger.debug('Creating Google VertexAI client');
+      return new GoogleVertexAI(clientOptions);
+    } else if (this.project_id && this.isChatModel) {
+      logger.debug('Creating Chat Google VertexAI client');
+      return new ChatGoogleVertexAI(clientOptions);
+    } else if (this.project_id) {
      logger.debug('Creating VertexAI client');
-      this.visionMode = undefined;
-      clientOptions.streaming = true;
-      const client = new ChatVertexAI(clientOptions);
-      client.temperature = clientOptions.temperature;
-      client.topP = clientOptions.topP;
-      client.topK = clientOptions.topK;
-      client.topLogprobs = clientOptions.topLogprobs;
-      client.frequencyPenalty = clientOptions.frequencyPenalty;
-      client.presencePenalty = clientOptions.presencePenalty;
-      client.maxOutputTokens = clientOptions.maxOutputTokens;
-      return client;
-    } else if (!EXCLUDED_GENAI_MODELS.test(model)) {
+      return new ChatVertexAI(clientOptions);
+    } else if (model.includes('1.5')) {
      logger.debug('Creating GenAI client');
-      return new GenAI(this.apiKey).getGenerativeModel({ model }, requestOptions);
+      return new GenAI(this.apiKey).getGenerativeModel(
+        {
+          ...clientOptions,
+          model,
+        },
+        { apiVersion: 'v1beta' },
+      );
    }

    logger.debug('Creating Chat Google Generative AI client');
    return new ChatGoogleGenerativeAI({ ...clientOptions, apiKey: this.apiKey });
  }

-  initializeClient() {
-    let clientOptions = { ...this.modelOptions };
+  async getCompletion(_payload, options = {}) {
+    const { parameters, instances } = _payload;
+    const { onProgress, abortController } = options;
+    const streamRate = this.options.streamRate ?? Constants.DEFAULT_STREAM_RATE;
+    const { messages: _messages, context, examples: _examples } = instances?.[0] ?? {};
+
+    let examples;
+
+    let clientOptions = { ...parameters, maxRetries: 2 };

    if (this.project_id) {
      clientOptions['authOptions'] = {
@@ -639,265 +638,185 @@ class GoogleClient extends BaseClient {
      };
    }

+    if (!parameters) {
+      clientOptions = { ...clientOptions, ...this.modelOptions };
+    }
+
    if (this.isGenerativeModel && !this.project_id) {
      clientOptions.modelName = clientOptions.model;
      delete clientOptions.model;
    }

-    this.client = this.createLLM(clientOptions);
-    return this.client;
-  }
+    if (_examples && _examples.length) {
+      examples = _examples
+        .map((ex) => {
+          const { input, output } = ex;
+          if (!input || !output) {
+            return undefined;
+          }
+          return {
+            input: new HumanMessage(input.content),
+            output: new AIMessage(output.content),
+          };
+        })
+        .filter((ex) => ex);

-  async getCompletion(_payload, options = {}) {
-    const { onProgress, abortController } = options;
-    const safetySettings = getSafetySettings(this.modelOptions.model);
-    const streamRate = this.options.streamRate ?? Constants.DEFAULT_STREAM_RATE;
-    const modelName = this.modelOptions.modelName ?? this.modelOptions.model ?? '';
+      clientOptions.examples = examples;
+    }
+
+    const model = this.createLLM(clientOptions);

    let reply = '';
-    /** @type {Error} */
-    let error;
-    try {
-      if (!EXCLUDED_GENAI_MODELS.test(modelName) && !this.project_id) {
-        /** @type {GenerativeModel} */
-        const client = this.client;
-        /** @type {GenerateContentRequest} */
-        const requestOptions = {
-          safetySettings,
-          contents: _payload,
-          generationConfig: googleGenConfigSchema.parse(this.modelOptions),
+    const messages = this.isTextModel ? _payload.trim() : _messages;
+
+    if (!this.isVisionModel && context && messages?.length > 0) {
+      messages.unshift(new SystemMessage(context));
+    }
+
+    const modelName = clientOptions.modelName ?? clientOptions.model ?? '';
+    if (modelName?.includes('1.5') && !this.project_id) {
+      const client = model;
+      const requestOptions = {
+        contents: _payload,
+      };
+
+      let promptPrefix = (this.options.promptPrefix ?? '').trim();
+      if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+        promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
+      }
+
+      if (this.options?.promptPrefix?.length) {
+        requestOptions.systemInstruction = {
+          parts: [
+            {
+              text: promptPrefix,
+            },
+          ],
        };
-
-        const promptPrefix = (this.systemMessage ?? '').trim();
-        if (promptPrefix.length) {
-          requestOptions.systemInstruction = {
-            parts: [
-              {
-                text: promptPrefix,
-              },
-            ],
-          };
-        }
-
-        const delay = modelName.includes('flash') ? 8 : 15;
-        /** @type {GenAIUsageMetadata} */
-        let usageMetadata;
-
-        abortController.signal.addEventListener(
-          'abort',
-          () => {
-            logger.warn('[GoogleClient] Request was aborted', abortController.signal.reason);
-          },
-          { once: true },
-        );
-
-        const result = await client.generateContentStream(requestOptions, {
-          signal: abortController.signal,
-        });
-        for await (const chunk of result.stream) {
-          usageMetadata = !usageMetadata
-            ? chunk?.usageMetadata
-            : Object.assign(usageMetadata, chunk?.usageMetadata);
-          const chunkText = chunk.text();
-          await this.generateTextStream(chunkText, onProgress, {
-            delay,
-          });
-          reply += chunkText;
-          await sleep(streamRate);
-        }
-
-        if (usageMetadata) {
-          this.usage = {
-            input_tokens: usageMetadata.promptTokenCount,
-            output_tokens: usageMetadata.candidatesTokenCount,
-          };
-        }
-
-        return reply;
      }

-      const { instances } = _payload;
-      const { messages: messages, context } = instances?.[0] ?? {};
+      requestOptions.safetySettings = _payload.safetySettings;

-      if (!this.isVisionModel && context && messages?.length > 0) {
-        messages.unshift(new SystemMessage(context));
-      }
-
-      /** @type {import('@langchain/core/messages').AIMessageChunk['usage_metadata']} */
-      let usageMetadata;
-      /** @type {ChatVertexAI} */
-      const client = this.client;
-      const stream = await client.stream(messages, {
-        signal: abortController.signal,
-        streamUsage: true,
-        safetySettings,
-      });
-
-      let delay = this.options.streamRate || 8;
-
-      if (!this.options.streamRate) {
-        if (this.isGenerativeModel) {
-          delay = 15;
-        }
-        if (modelName.includes('flash')) {
-          delay = 5;
-        }
-      }
-
-      for await (const chunk of stream) {
-        if (chunk?.usage_metadata) {
-          const metadata = chunk.usage_metadata;
-          for (const key in metadata) {
-            if (Number.isNaN(metadata[key])) {
-              delete metadata[key];
-            }
-          }
-
-          usageMetadata = !usageMetadata ? metadata : concat(usageMetadata, metadata);
-        }
-
-        const chunkText = chunk?.content ?? '';
+      const delay = modelName.includes('flash') ? 8 : 14;
+      const result = await client.generateContentStream(requestOptions);
+      for await (const chunk of result.stream) {
+        const chunkText = chunk.text();
        await this.generateTextStream(chunkText, onProgress, {
          delay,
        });
        reply += chunkText;
+        await sleep(streamRate);
      }
-
-      if (usageMetadata) {
-        this.usage = usageMetadata;
-      }
-    } catch (e) {
-      error = e;
-      logger.error('[GoogleClient] There was an issue generating the completion', e);
+      return reply;
    }

-    if (error != null && reply === '') {
-      const errorMessage = `{ "type": "${ErrorTypes.GoogleError}", "info": "${
-        error.message ?? 'The Google provider failed to generate content, please contact the Admin.'
-      }" }`;
-      throw new Error(errorMessage);
+    const stream = await model.stream(messages, {
+      signal: abortController.signal,
+      timeout: 7000,
+      safetySettings: _payload.safetySettings,
+    });
+
+    let delay = this.options.streamRate || 8;
+
+    if (!this.options.streamRate) {
+      if (this.isGenerativeModel) {
+        delay = 12;
+      }
+      if (modelName.includes('flash')) {
+        delay = 5;
+      }
    }
+
+    for await (const chunk of stream) {
+      const chunkText = chunk?.content ?? chunk;
+      await this.generateTextStream(chunkText, onProgress, {
+        delay,
+      });
+      reply += chunkText;
+    }
+
    return reply;
  }

-  /**
-   * Get stream usage as returned by this client's API response.
-   * @returns {UsageMetadata} The stream usage object.
-   */
-  getStreamUsage() {
-    return this.usage;
-  }
-
-  getMessageMapMethod() {
-    /**
-     * @param {TMessage} msg
-     */
-    return (msg) => {
-      if (msg.text != null && msg.text && msg.text.startsWith(':::thinking')) {
-        msg.text = msg.text.replace(/:::thinking.*?:::/gs, '').trim();
-      } else if (msg.content != null) {
-        msg.text = parseTextParts(msg.content, true);
-        delete msg.content;
-      }
-
-      return msg;
-    };
-  }
-
-  /**
-   * Calculates the correct token count for the current user message based on the token count map and API usage.
-   * Edge case: If the calculation results in a negative value, it returns the original estimate.
-   * If revisiting a conversation with a chat history entirely composed of token estimates,
-   * the cumulative token count going forward should become more accurate as the conversation progresses.
-   * @param {Object} params - The parameters for the calculation.
-   * @param {Record<string, number>} params.tokenCountMap - A map of message IDs to their token counts.
-   * @param {string} params.currentMessageId - The ID of the current message to calculate.
-   * @param {UsageMetadata} params.usage - The usage object returned by the API.
-   * @returns {number} The correct token count for the current user message.
-   */
-  calculateCurrentTokenCount({ tokenCountMap, currentMessageId, usage }) {
-    const originalEstimate = tokenCountMap[currentMessageId] || 0;
-
-    if (!usage || typeof usage.input_tokens !== 'number') {
-      return originalEstimate;
-    }
-
-    tokenCountMap[currentMessageId] = 0;
-    const totalTokensFromMap = Object.values(tokenCountMap).reduce((sum, count) => {
-      const numCount = Number(count);
-      return sum + (isNaN(numCount) ? 0 : numCount);
-    }, 0);
-    const totalInputTokens = usage.input_tokens ?? 0;
-    const currentMessageTokens = totalInputTokens - totalTokensFromMap;
-    return currentMessageTokens > 0 ? currentMessageTokens : originalEstimate;
-  }
-
-  /**
-   * @param {object} params
-   * @param {number} params.promptTokens
-   * @param {number} params.completionTokens
-   * @param {UsageMetadata} [params.usage]
-   * @param {string} [params.model]
-   * @param {string} [params.context='message']
-   * @returns {Promise<void>}
-   */
-  async recordTokenUsage({ promptTokens, completionTokens, model, context = 'message' }) {
-    await spendTokens(
-      {
-        context,
-        user: this.user ?? this.options.req?.user?.id,
-        conversationId: this.conversationId,
-        model: model ?? this.modelOptions.model,
-        endpointTokenConfig: this.options.endpointTokenConfig,
-      },
-      { promptTokens, completionTokens },
-    );
-  }
-
  /**
   * Stripped-down logic for generating a title. This uses the non-streaming APIs, since the user does not see titles streaming
   */
  async titleChatCompletion(_payload, options = {}) {
-    let reply = '';
    const { abortController } = options;
+    const { parameters, instances } = _payload;
+    const { messages: _messages, examples: _examples } = instances?.[0] ?? {};

-    const model =
-      this.options.titleModel ?? this.modelOptions.modelName ?? this.modelOptions.model ?? '';
-    const safetySettings = getSafetySettings(model);
-    if (!EXCLUDED_GENAI_MODELS.test(model) && !this.project_id) {
-      logger.debug('Identified titling model as GenAI version');
+    let clientOptions = { ...parameters, maxRetries: 2 };
+
+    logger.debug('Initialized title client options');
+
+    if (this.project_id) {
+      clientOptions['authOptions'] = {
+        credentials: {
+          ...this.serviceKey,
+        },
+        projectId: this.project_id,
+      };
+    }
+
+    if (!parameters) {
+      clientOptions = { ...clientOptions, ...this.modelOptions };
+    }
+
+    if (this.isGenerativeModel && !this.project_id) {
+      clientOptions.modelName = clientOptions.model;
+      delete clientOptions.model;
+    }
+
+    const model = this.createLLM(clientOptions);
+
+    let reply = '';
+    const messages = this.isTextModel ? _payload.trim() : _messages;
+
+    const modelName = clientOptions.modelName ?? clientOptions.model ?? '';
+    if (modelName?.includes('1.5') && !this.project_id) {
+      logger.debug('Identified titling model as 1.5 version');
      /** @type {GenerativeModel} */
-      const client = this.client;
+      const client = model;
      const requestOptions = {
        contents: _payload,
-        safetySettings,
-        generationConfig: {
-          temperature: 0.5,
-        },
      };

-      const result = await client.generateContent(requestOptions);
-      reply = result.response?.text();
-      return reply;
-    } else {
-      const { instances } = _payload;
-      const { messages } = instances?.[0] ?? {};
-      const titleResponse = await this.client.invoke(messages, {
-        signal: abortController.signal,
-        timeout: 7000,
-        safetySettings,
-      });
-
-      if (titleResponse.usage_metadata) {
-        await this.recordTokenUsage({
-          model,
-          promptTokens: titleResponse.usage_metadata.input_tokens,
-          completionTokens: titleResponse.usage_metadata.output_tokens,
-          context: 'title',
-        });
+      let promptPrefix = (this.options.promptPrefix ?? '').trim();
+      if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+        promptPrefix = `${promptPrefix ?? ''}\n${this.options.artifactsPrompt}`.trim();
      }

+      if (this.options?.promptPrefix?.length) {
+        requestOptions.systemInstruction = {
+          parts: [
+            {
+              text: promptPrefix,
+            },
+          ],
+        };
+      }
+
+      const safetySettings = _payload.safetySettings;
+      requestOptions.safetySettings = safetySettings;
+
+      const result = await client.generateContent(requestOptions);
+
+      reply = result.response?.text();
+
+      return reply;
+    } else {
+      logger.debug('Beginning titling');
+      const safetySettings = _payload.safetySettings;
+
+      const titleResponse = await model.invoke(messages, {
+        signal: abortController.signal,
+        timeout: 7000,
+        safetySettings: safetySettings,
+      });
+
      reply = titleResponse.content;
+      // TODO: RECORD TOKEN USAGE
      return reply;
    }
  }
@@ -921,8 +840,15 @@ class GoogleClient extends BaseClient {
      },
    ]);

+    if (this.isVisionModel) {
+      logger.warn(
+        `Current vision model does not support titling without an attachment; falling back to default model ${settings.model.default}`,
+      );
+
+      payload.parameters = { ...payload.parameters, model: settings.model.default };
+    }
+
    try {
-      this.initializeClient();
      title = await this.titleChatCompletion(payload, {
        abortController: new AbortController(),
        onProgress: () => {},
@@ -936,10 +862,8 @@ class GoogleClient extends BaseClient {

  getSaveOptions() {
    return {
-      endpointType: null,
      artifacts: this.options.artifacts,
      promptPrefix: this.options.promptPrefix,
-      maxContextTokens: this.options.maxContextTokens,
      modelLabel: this.options.modelLabel,
      iconURL: this.options.iconURL,
      greeting: this.options.greeting,
@@ -953,39 +877,53 @@ class GoogleClient extends BaseClient {
  }

  async sendCompletion(payload, opts = {}) {
+    payload.safetySettings = this.getSafetySettings();
+
    let reply = '';
    reply = await this.getCompletion(payload, opts);
    return reply.trim();
  }

-  getEncoding() {
-    return 'cl100k_base';
-  }
-
-  async getVertexTokenCount(text) {
-    /** @type {ChatVertexAI} */
-    const client = this.client ?? this.initializeClient();
-    const connection = client.connection;
-    const gAuthClient = connection.client;
-    const tokenEndpoint = `https://${connection._endpoint}/${connection.apiVersion}/projects/${this.project_id}/locations/${connection._location}/publishers/google/models/${connection.model}/:countTokens`;
-    const result = await gAuthClient.request({
-      url: tokenEndpoint,
-      method: 'POST',
-      data: {
-        contents: [{ role: 'user', parts: [{ text }] }],
+  getSafetySettings() {
+    return [
+      {
+        category: 'HARM_CATEGORY_SEXUALLY_EXPLICIT',
+        threshold:
+          process.env.GOOGLE_SAFETY_SEXUALLY_EXPLICIT || 'HARM_BLOCK_THRESHOLD_UNSPECIFIED',
      },
-    });
-    return result;
+      {
+        category: 'HARM_CATEGORY_HATE_SPEECH',
+        threshold: process.env.GOOGLE_SAFETY_HATE_SPEECH || 'HARM_BLOCK_THRESHOLD_UNSPECIFIED',
+      },
+      {
+        category: 'HARM_CATEGORY_HARASSMENT',
+        threshold: process.env.GOOGLE_SAFETY_HARASSMENT || 'HARM_BLOCK_THRESHOLD_UNSPECIFIED',
+      },
+      {
+        category: 'HARM_CATEGORY_DANGEROUS_CONTENT',
+        threshold:
+          process.env.GOOGLE_SAFETY_DANGEROUS_CONTENT || 'HARM_BLOCK_THRESHOLD_UNSPECIFIED',
+      },
+    ];
+  }
+
+  /* TO-DO: Handle tokens with Google tokenization NOTE: these are required */
+  static getTokenizer(encoding, isModelName = false, extendSpecialTokens = {}) {
+    if (tokenizersCache[encoding]) {
+      return tokenizersCache[encoding];
+    }
+    let tokenizer;
+    if (isModelName) {
+      tokenizer = encodingForModel(encoding, extendSpecialTokens);
+    } else {
+      tokenizer = getEncoding(encoding, extendSpecialTokens);
+    }
+    tokenizersCache[encoding] = tokenizer;
+    return tokenizer;
  }

-  /**
-   * Returns the token count of a given text. It also checks and resets the tokenizers if necessary.
-   * @param {string} text - The text to get the token count for.
-   * @returns {number} The token count of the given text.
-   */
  getTokenCount(text) {
-    const encoding = this.getEncoding();
-    return Tokenizer.getTokenCount(text, encoding);
+    return this.gptEncoder.encode(text, 'all').length;
  }
 }

--- a/api/app/clients/OllamaClient.js
+++ b/api/app/clients/OllamaClient.js
@@ -1,11 +1,10 @@
 const { z } = require('zod');
 const axios = require('axios');
 const { Ollama } = require('ollama');
-const { sleep } = require('@librechat/agents');
-const { logAxiosError } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { Constants } = require('librechat-data-provider');
 const { deriveBaseURL } = require('~/utils');
+const { sleep } = require('~/server/utils');
+const { logger } = require('~/config');

 const ollamaPayloadSchema = z.object({
  mirostat: z.number().optional(),
@@ -68,8 +67,8 @@ class OllamaClient {
      return models;
    } catch (error) {
      const logMessage =
-        "Failed to fetch models from Ollama API. If you are not using Ollama directly, and instead, through some aggregator or reverse proxy that handles fetching via OpenAI spec, ensure the name of the endpoint doesn't start with `ollama` (case-insensitive).";
-      logAxiosError({ message: logMessage, error });
+        'Failed to fetch models from Ollama API. If you are not using Ollama directly, and instead, through some aggregator or reverse proxy that handles fetching via OpenAI spec, ensure the name of the endpoint doesn\'t start with `ollama` (case-insensitive).';
+      logger.error(logMessage, error);
      return [];
    }
  }
--- a/api/app/clients/OpenAIClient.js
+++ b/api/app/clients/OpenAIClient.js
@@ -1,24 +1,11 @@
+const OpenAI = require('openai');
 const { OllamaClient } = require('./OllamaClient');
 const { HttpsProxyAgent } = require('https-proxy-agent');
-const { SplitStreamHandler, CustomOpenAIClient: OpenAI } = require('@librechat/agents');
-const {
-  isEnabled,
-  Tokenizer,
-  createFetch,
-  resolveHeaders,
-  constructAzureURL,
-  getModelMaxTokens,
-  genAzureChatCompletion,
-  getModelMaxOutputTokens,
-  createStreamEventHandlers,
-} = require('@librechat/api');
 const {
  Constants,
  ImageDetail,
-  ContentTypes,
-  parseTextParts,
  EModelEndpoint,
-  KnownEndpoints,
+  resolveHeaders,
  openAISettings,
  ImageDetailCost,
  CohereConstants,
@@ -26,6 +13,14 @@ const {
  validateVisionModel,
  mapModelToAzureConfig,
 } = require('librechat-data-provider');
+const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
+const {
+  extractBaseURL,
+  constructAzureURL,
+  getModelMaxTokens,
+  genAzureChatCompletion,
+  getModelMaxOutputTokens,
+} = require('~/utils');
 const {
  truncateText,
  formatMessage,
@@ -34,20 +29,31 @@ const {
  createContextHandlers,
 } = require('./prompts');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const { addSpaceIfNeeded, sleep } = require('~/server/utils');
 const { spendTokens } = require('~/models/spendTokens');
+const { isEnabled, sleep } = require('~/server/utils');
 const { handleOpenAIErrors } = require('./tools/util');
+const { createLLM, RunManager } = require('./llm');
+const ChatGPTClient = require('./ChatGPTClient');
 const { summaryBuffer } = require('./memory');
 const { runTitleChain } = require('./chains');
-const { extractBaseURL } = require('~/utils');
 const { tokenSplit } = require('./document');
 const BaseClient = require('./BaseClient');
-const { createLLM } = require('./llm');
 const { logger } = require('~/config');

+// Cache to store Tiktoken instances
+const tokenizersCache = {};
+// Counter for keeping track of the number of tokenizer calls
+let tokenizerCallsCount = 0;
+
 class OpenAIClient extends BaseClient {
  constructor(apiKey, options = {}) {
    super(apiKey, options);
+    this.ChatGPTClient = new ChatGPTClient();
+    this.buildPrompt = this.ChatGPTClient.buildPrompt.bind(this);
+    /** @type {getCompletion} */
+    this.getCompletion = this.ChatGPTClient.getCompletion.bind(this);
+    /** @type {cohereChatCompletion} */
+    this.cohereChatCompletion = this.ChatGPTClient.cohereChatCompletion.bind(this);
    this.contextStrategy = options.contextStrategy
      ? options.contextStrategy.toLowerCase()
      : 'discard';
@@ -63,9 +69,7 @@ class OpenAIClient extends BaseClient {
    /** @type {OpenAIUsageMetadata | undefined} */
    this.usage;
    /** @type {boolean|undefined} */
-    this.isOmni;
-    /** @type {SplitStreamHandler | undefined} */
-    this.streamHandler;
+    this.isO1Model;
  }

  // TODO: PluginsClient calls this 3x, unneeded
@@ -103,17 +107,20 @@ class OpenAIClient extends BaseClient {
      this.checkVisionRequest(this.options.attachments);
    }

-    const omniPattern = /\b(o\d)\b/i;
-    this.isOmni = omniPattern.test(this.modelOptions.model);
+    this.isO1Model = /\bo1\b/i.test(this.modelOptions.model);
+
+    const { OPENROUTER_API_KEY, OPENAI_FORCE_PROMPT } = process.env ?? {};
+    if (OPENROUTER_API_KEY && !this.azure) {
+      this.apiKey = OPENROUTER_API_KEY;
+      this.useOpenRouter = true;
+    }

-    const { OPENAI_FORCE_PROMPT } = process.env ?? {};
    const { reverseProxyUrl: reverseProxy } = this.options;

    if (
      !this.useOpenRouter &&
-      ((reverseProxy && reverseProxy.includes(KnownEndpoints.openrouter)) ||
-        (this.options.endpoint &&
-          this.options.endpoint.toLowerCase().includes(KnownEndpoints.openrouter)))
+      reverseProxy &&
+      reverseProxy.includes('https://openrouter.ai/api/v1')
    ) {
      this.useOpenRouter = true;
    }
@@ -140,7 +147,7 @@ class OpenAIClient extends BaseClient {
    const { model } = this.modelOptions;

    this.isChatCompletion =
-      omniPattern.test(model) || model.includes('gpt') || this.useOpenRouter || !!reverseProxy;
+      /\bo1\b/i.test(model) || model.includes('gpt') || this.useOpenRouter || !!reverseProxy;
    this.isChatGptModel = this.isChatCompletion;
    if (
      model.includes('text-davinci') ||
@@ -222,6 +229,10 @@ class OpenAIClient extends BaseClient {
      logger.debug('Using Azure endpoint');
    }

+    if (this.useOpenRouter) {
+      this.completionsUrl = 'https://openrouter.ai/api/v1/chat/completions';
+    }
+
    return this;
  }

@@ -295,10 +306,75 @@ class OpenAIClient extends BaseClient {
    }
  }

-  getEncoding() {
-    return this.modelOptions?.model && /gpt-4[^-\s]/.test(this.modelOptions.model)
-      ? 'o200k_base'
-      : 'cl100k_base';
+  // Selects an appropriate tokenizer based on the current configuration of the client instance.
+  // It takes into account factors such as whether it's a chat completion, an unofficial chat GPT model, etc.
+  selectTokenizer() {
+    let tokenizer;
+    this.encoding = 'text-davinci-003';
+    if (this.isChatCompletion) {
+      this.encoding = this.modelOptions.model.includes('gpt-4o') ? 'o200k_base' : 'cl100k_base';
+      tokenizer = this.constructor.getTokenizer(this.encoding);
+    } else if (this.isUnofficialChatGptModel) {
+      const extendSpecialTokens = {
+        '<|im_start|>': 100264,
+        '<|im_end|>': 100265,
+      };
+      tokenizer = this.constructor.getTokenizer(this.encoding, true, extendSpecialTokens);
+    } else {
+      try {
+        const { model } = this.modelOptions;
+        this.encoding = model.includes('instruct') ? 'text-davinci-003' : model;
+        tokenizer = this.constructor.getTokenizer(this.encoding, true);
+      } catch {
+        tokenizer = this.constructor.getTokenizer('text-davinci-003', true);
+      }
+    }
+
+    return tokenizer;
+  }
+
+  // Retrieves a tokenizer either from the cache or creates a new one if one doesn't exist in the cache.
+  // If a tokenizer is being created, it's also added to the cache.
+  static getTokenizer(encoding, isModelName = false, extendSpecialTokens = {}) {
+    let tokenizer;
+    if (tokenizersCache[encoding]) {
+      tokenizer = tokenizersCache[encoding];
+    } else {
+      if (isModelName) {
+        tokenizer = encodingForModel(encoding, extendSpecialTokens);
+      } else {
+        tokenizer = getEncoding(encoding, extendSpecialTokens);
+      }
+      tokenizersCache[encoding] = tokenizer;
+    }
+    return tokenizer;
+  }
+
+  // Frees all encoders in the cache and resets the count.
+  static freeAndResetAllEncoders() {
+    try {
+      Object.keys(tokenizersCache).forEach((key) => {
+        if (tokenizersCache[key]) {
+          tokenizersCache[key].free();
+          delete tokenizersCache[key];
+        }
+      });
+      // Reset count
+      tokenizerCallsCount = 1;
+    } catch (error) {
+      logger.error('[OpenAIClient] Free and reset encoders error', error);
+    }
+  }
+
+  // Checks if the cache of tokenizers has reached a certain size. If it has, it frees and resets all tokenizers.
+  resetTokenizersIfNecessary() {
+    if (tokenizerCallsCount >= 25) {
+      if (this.options.debug) {
+        logger.debug('[OpenAIClient] freeAndResetAllEncoders: reached 25 encodings, resetting...');
+      }
+      this.constructor.freeAndResetAllEncoders();
+    }
+    tokenizerCallsCount++;
  }

  /**
@@ -307,8 +383,15 @@ class OpenAIClient extends BaseClient {
   * @returns {number} The token count of the given text.
   */
  getTokenCount(text) {
-    const encoding = this.getEncoding();
-    return Tokenizer.getTokenCount(text, encoding);
+    this.resetTokenizersIfNecessary();
+    try {
+      const tokenizer = this.selectTokenizer();
+      return tokenizer.encode(text, 'all').length;
+    } catch (error) {
+      this.constructor.freeAndResetAllEncoders();
+      const tokenizer = this.selectTokenizer();
+      return tokenizer.encode(text, 'all').length;
+    }
  }

  /**
@@ -340,7 +423,6 @@ class OpenAIClient extends BaseClient {
      promptPrefix: this.options.promptPrefix,
      resendFiles: this.options.resendFiles,
      imageDetail: this.options.imageDetail,
-      modelLabel: this.options.modelLabel,
      iconURL: this.options.iconURL,
      greeting: this.options.greeting,
      spec: this.options.spec,
@@ -374,12 +456,23 @@ class OpenAIClient extends BaseClient {
    return files;
  }

-  async buildMessages(messages, parentMessageId, { promptPrefix = null }, opts) {
+  async buildMessages(
+    messages,
+    parentMessageId,
+    { isChatCompletion = false, promptPrefix = null },
+    opts,
+  ) {
    let orderedMessages = this.constructor.getMessagesForConversation({
      messages,
      parentMessageId,
      summary: this.shouldSummarize,
    });
+    if (!isChatCompletion) {
+      return await this.buildPrompt(orderedMessages, {
+        isChatGptModel: isChatCompletion,
+        promptPrefix,
+      });
+    }

    let payload;
    let instructions;
@@ -439,9 +532,6 @@ class OpenAIClient extends BaseClient {
            this.contextHandlers?.processFile(file);
            continue;
          }
-          if (file.metadata?.fileIdentifier) {
-            continue;
-          }

          orderedMessages[i].tokenCount += this.calculateImageTokenCost({
            width: file.width,
@@ -459,9 +549,7 @@ class OpenAIClient extends BaseClient {
      promptPrefix = this.augmentedPrompt + promptPrefix;
    }

-    const noSystemModelRegex = /\b(o1-preview|o1-mini)\b/i.test(this.modelOptions.model);
-
-    if (promptPrefix && !noSystemModelRegex) {
+    if (promptPrefix && this.isO1Model !== true) {
      promptPrefix = `Instructions:\n${promptPrefix.trim()}`;
      instructions = {
        role: 'system',
@@ -489,27 +577,12 @@ class OpenAIClient extends BaseClient {
    };

    /** EXPERIMENTAL */
-    if (promptPrefix && noSystemModelRegex) {
+    if (promptPrefix && this.isO1Model === true) {
      const lastUserMessageIndex = payload.findLastIndex((message) => message.role === 'user');
      if (lastUserMessageIndex !== -1) {
-        if (Array.isArray(payload[lastUserMessageIndex].content)) {
-          const firstTextPartIndex = payload[lastUserMessageIndex].content.findIndex(
-            (part) => part.type === ContentTypes.TEXT,
-          );
-          if (firstTextPartIndex !== -1) {
-            const firstTextPart = payload[lastUserMessageIndex].content[firstTextPartIndex];
-            payload[lastUserMessageIndex].content[firstTextPartIndex].text =
-              `${promptPrefix}\n${firstTextPart.text}`;
-          } else {
-            payload[lastUserMessageIndex].content.unshift({
-              type: ContentTypes.TEXT,
-              text: promptPrefix,
-            });
-          }
-        } else {
-          payload[lastUserMessageIndex].content =
-            `${promptPrefix}\n${payload[lastUserMessageIndex].content}`;
-        }
+        payload[
+          lastUserMessageIndex
+        ].content = `${promptPrefix}\n${payload[lastUserMessageIndex].content}`;
      }
    }

@@ -598,7 +671,7 @@ class OpenAIClient extends BaseClient {
        return result.trim();
      }

-      logger.debug('[OpenAIClient] sendCompletion: result', { ...result });
+      logger.debug('[OpenAIClient] sendCompletion: result', result);

      if (this.isChatCompletion) {
        reply = result.choices[0].message.content;
@@ -615,15 +688,23 @@ class OpenAIClient extends BaseClient {
  }

  initializeLLM({
-    model = openAISettings.model.default,
+    model = 'gpt-3.5-turbo',
    modelName,
    temperature = 0.2,
+    presence_penalty = 0,
+    frequency_penalty = 0,
    max_tokens,
    streaming,
+    context,
+    tokenBuffer,
+    initialMessageCount,
+    conversationId,
  }) {
    const modelOptions = {
      modelName: modelName ?? model,
      temperature,
+      presence_penalty,
+      frequency_penalty,
      user: this.user,
    };

@@ -651,10 +732,8 @@ class OpenAIClient extends BaseClient {
    if (headers && typeof headers === 'object' && !Array.isArray(headers)) {
      configOptions.baseOptions = {
        headers: resolveHeaders({
-          headers: {
-            ...headers,
-            ...configOptions?.baseOptions?.headers,
-          },
+          ...headers,
+          ...configOptions?.baseOptions?.headers,
        }),
      };
    }
@@ -664,12 +743,22 @@ class OpenAIClient extends BaseClient {
      configOptions.httpsAgent = new HttpsProxyAgent(this.options.proxy);
    }

+    const { req, res, debug } = this.options;
+    const runManager = new RunManager({ req, res, debug, abortController: this.abortController });
+    this.runManager = runManager;
+
    const llm = createLLM({
      modelOptions,
      configOptions,
      openAIApiKey: this.apiKey,
      azure: this.azure,
      streaming,
+      callbacks: runManager.createCallbacks({
+        context,
+        tokenBuffer,
+        conversationId: this.conversationId ?? conversationId,
+        initialMessageCount,
+      }),
    });

    return llm;
@@ -690,7 +779,6 @@ class OpenAIClient extends BaseClient {
   *                            In case of failure, it will return the default title, "New Chat".
   */
  async titleConvo({ text, conversationId, responseText = '' }) {
-    const appConfig = this.options.req?.config;
    this.conversationId = conversationId;

    if (this.options.attachments) {
@@ -705,7 +793,7 @@ class OpenAIClient extends BaseClient {

    const { OPENAI_TITLE_MODEL } = process.env ?? {};

-    let model = this.options.titleModel ?? OPENAI_TITLE_MODEL ?? openAISettings.model.default;
+    let model = this.options.titleModel ?? OPENAI_TITLE_MODEL ?? 'gpt-3.5-turbo';
    if (model === Constants.CURRENT_MODEL) {
      model = this.modelOptions.model;
    }
@@ -719,7 +807,8 @@ class OpenAIClient extends BaseClient {
      max_tokens: 16,
    };

-    const azureConfig = appConfig?.endpoints?.[EModelEndpoint.azureOpenAI];
+    /** @type {TAzureConfig | undefined} */
+    const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];

    const resetTitleOptions = !!(
      (this.azure && azureConfig) ||
@@ -739,7 +828,7 @@ class OpenAIClient extends BaseClient {
        groupMap,
      });

-      this.options.headers = resolveHeaders({ headers });
+      this.options.headers = resolveHeaders(headers);
      this.options.reverseProxyUrl = baseURL ?? null;
      this.langchainProxy = extractBaseURL(this.options.reverseProxyUrl);
      this.apiKey = azureOptions.azureOpenAIApiKey;
@@ -749,12 +838,6 @@ class OpenAIClient extends BaseClient {
      this.options.dropParams = azureConfig.groupMap[groupName].dropParams;
      this.options.forcePrompt = azureConfig.groupMap[groupName].forcePrompt;
      this.azure = !serverless && azureOptions;
-      if (serverless === true) {
-        this.options.defaultQuery = azureOptions.azureOpenAIApiVersion
-          ? { 'api-version': azureOptions.azureOpenAIApiVersion }
-          : undefined;
-        this.options.headers['api-key'] = this.apiKey;
-      }
    }

    const titleChatCompletion = async () => {
@@ -786,16 +869,12 @@ ${convo}
        }

        title = (
-          await this.sendPayload(instructionsPayload, {
-            modelOptions,
-            useChatCompletion,
-            context: 'title',
-          })
+          await this.sendPayload(instructionsPayload, { modelOptions, useChatCompletion })
        ).replaceAll('"', '');

        const completionTokens = this.getTokenCount(title);

-        await this.recordTokenUsage({ promptTokens, completionTokens, context: 'title' });
+        this.recordTokenUsage({ promptTokens, completionTokens, context: 'title' });
      } catch (e) {
        logger.error(
          '[OpenAIClient] There was an issue generating the title with the completion method',
@@ -897,7 +976,7 @@ ${convo}
    let prompt;

    // TODO: remove the gpt fallback and make it specific to endpoint
-    const { OPENAI_SUMMARY_MODEL = openAISettings.model.default } = process.env ?? {};
+    const { OPENAI_SUMMARY_MODEL = 'gpt-3.5-turbo' } = process.env ?? {};
    let model = this.options.summaryModel ?? OPENAI_SUMMARY_MODEL;
    if (model === Constants.CURRENT_MODEL) {
      model = this.modelOptions.model;
@@ -923,10 +1002,7 @@ ${convo}
    );

    if (excessTokenCount > maxContextTokens) {
-      ({ context } = await this.getMessagesWithinTokenLimit({
-        messages: context,
-        maxContextTokens,
-      }));
+      ({ context } = await this.getMessagesWithinTokenLimit(context, maxContextTokens));
    }

    if (context.length === 0) {
@@ -1056,62 +1132,10 @@ ${convo}
    });
  }

-  /**
-   *
-   * @param {string[]} [intermediateReply]
-   * @returns {string}
-   */
-  getStreamText(intermediateReply) {
-    if (!this.streamHandler) {
-      return intermediateReply?.join('') ?? '';
-    }
-
-    let thinkMatch;
-    let remainingText;
-    let reasoningText = '';
-
-    if (this.streamHandler.reasoningTokens.length > 0) {
-      reasoningText = this.streamHandler.reasoningTokens.join('');
-      thinkMatch = reasoningText.match(/<think>([\s\S]*?)<\/think>/)?.[1]?.trim();
-      if (thinkMatch != null && thinkMatch) {
-        const reasoningTokens = `:::thinking\n${thinkMatch}\n:::\n`;
-        remainingText = reasoningText.split(/<\/think>/)?.[1]?.trim() || '';
-        return `${reasoningTokens}${remainingText}${this.streamHandler.tokens.join('')}`;
-      } else if (thinkMatch === '') {
-        remainingText = reasoningText.split(/<\/think>/)?.[1]?.trim() || '';
-        return `${remainingText}${this.streamHandler.tokens.join('')}`;
-      }
-    }
-
-    const reasoningTokens =
-      reasoningText.length > 0
-        ? `:::thinking\n${reasoningText.replace('<think>', '').replace('</think>', '').trim()}\n:::\n`
-        : '';
-
-    return `${reasoningTokens}${this.streamHandler.tokens.join('')}`;
-  }
-
-  getMessageMapMethod() {
-    /**
-     * @param {TMessage} msg
-     */
-    return (msg) => {
-      if (msg.text != null && msg.text && msg.text.startsWith(':::thinking')) {
-        msg.text = msg.text.replace(/:::thinking.*?:::/gs, '').trim();
-      } else if (msg.content != null) {
-        msg.text = parseTextParts(msg.content, true);
-        delete msg.content;
-      }
-
-      return msg;
-    };
-  }
-
  async chatCompletion({ payload, onProgress, abortController = null }) {
-    const appConfig = this.options.req?.config;
    let error = null;
-    let intermediateReply = [];
    const errorCallback = (err) => (error = err);
+    const intermediateReply = [];
    try {
      if (!abortController) {
        abortController = new AbortController();
@@ -1132,7 +1156,6 @@ ${convo}
      logger.debug('[OpenAIClient] chatCompletion', { baseURL, modelOptions });
      const opts = {
        baseURL,
-        fetchOptions: {},
      };

      if (this.useOpenRouter) {
@@ -1146,15 +1169,16 @@ ${convo}
        opts.defaultHeaders = { ...opts.defaultHeaders, ...this.options.headers };
      }

-      if (this.options.defaultQuery) {
-        opts.defaultQuery = this.options.defaultQuery;
-      }
-
      if (this.options.proxy) {
-        opts.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
+        opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
      }

-      const azureConfig = appConfig?.endpoints?.[EModelEndpoint.azureOpenAI];
+      if (this.isVisionModel) {
+        modelOptions.max_tokens = 4000;
+      }
+
+      /** @type {TAzureConfig | undefined} */
+      const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];

      if (
        (this.azure && this.isVisionModel && azureConfig) ||
@@ -1171,7 +1195,7 @@ ${convo}
          modelGroupMap,
          groupMap,
        });
-        opts.defaultHeaders = resolveHeaders({ headers });
+        opts.defaultHeaders = resolveHeaders(headers);
        this.langchainProxy = extractBaseURL(baseURL);
        this.apiKey = azureOptions.azureOpenAIApiKey;

@@ -1183,12 +1207,6 @@ ${convo}
        this.azure = !serverless && azureOptions;
        this.azureEndpoint =
          !serverless && genAzureChatCompletion(this.azure, modelOptions.model, this);
-        if (serverless === true) {
-          this.options.defaultQuery = azureOptions.azureOpenAIApiVersion
-            ? { 'api-version': azureOptions.azureOpenAIApiVersion }
-            : undefined;
-          this.options.headers['api-key'] = this.apiKey;
-        }
      }

      if (this.azure || this.options.azure) {
@@ -1202,24 +1220,19 @@ ${convo}

        opts.baseURL = this.langchainProxy
          ? constructAzureURL({
-              baseURL: this.langchainProxy,
-              azureOptions: this.azure,
-            })
+            baseURL: this.langchainProxy,
+            azureOptions: this.azure,
+          })
          : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];

        opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
        opts.defaultHeaders = { ...opts.defaultHeaders, 'api-key': this.apiKey };
      }

-      if (this.isOmni === true && modelOptions.max_tokens != null) {
-        const paramName =
-          modelOptions.useResponsesApi === true ? 'max_output_tokens' : 'max_completion_tokens';
-        modelOptions[paramName] = modelOptions.max_tokens;
+      if (this.isO1Model === true && modelOptions.max_tokens != null) {
+        modelOptions.max_completion_tokens = modelOptions.max_tokens;
        delete modelOptions.max_tokens;
      }
-      if (this.isOmni === true && modelOptions.temperature != null) {
-        delete modelOptions.temperature;
-      }

      if (process.env.OPENAI_ORGANIZATION) {
        opts.organization = process.env.OPENAI_ORGANIZATION;
@@ -1228,10 +1241,7 @@ ${convo}
      let chatCompletion;
      /** @type {OpenAI} */
      const openai = new OpenAI({
-        fetch: createFetch({
-          directEndpoint: this.options.directEndpoint,
-          reverseProxyUrl: this.options.reverseProxyUrl,
-        }),
+        fetch: this.fetch,
        apiKey: this.apiKey,
        ...opts,
      });
@@ -1260,56 +1270,23 @@ ${convo}
        modelOptions.messages[0].role = 'user';
      }

-      if (
-        (this.options.endpoint === EModelEndpoint.openAI ||
-          this.options.endpoint === EModelEndpoint.azureOpenAI) &&
-        modelOptions.stream === true
-      ) {
-        modelOptions.stream_options = { include_usage: true };
-      }
-
      if (this.options.addParams && typeof this.options.addParams === 'object') {
-        const addParams = { ...this.options.addParams };
        modelOptions = {
          ...modelOptions,
-          ...addParams,
+          ...this.options.addParams,
        };
        logger.debug('[OpenAIClient] chatCompletion: added params', {
-          addParams: addParams,
+          addParams: this.options.addParams,
          modelOptions,
        });
      }

-      /** Note: OpenAI Web Search models do not support any known parameters besdies `max_tokens` */
-      if (modelOptions.model && /gpt-4o.*search/.test(modelOptions.model)) {
-        const searchExcludeParams = [
-          'frequency_penalty',
-          'presence_penalty',
-          'temperature',
-          'top_p',
-          'top_k',
-          'stop',
-          'logit_bias',
-          'seed',
-          'response_format',
-          'n',
-          'logprobs',
-          'user',
-        ];
-
-        this.options.dropParams = this.options.dropParams || [];
-        this.options.dropParams = [
-          ...new Set([...this.options.dropParams, ...searchExcludeParams]),
-        ];
-      }
-
      if (this.options.dropParams && Array.isArray(this.options.dropParams)) {
-        const dropParams = [...this.options.dropParams];
-        dropParams.forEach((param) => {
+        this.options.dropParams.forEach((param) => {
          delete modelOptions[param];
        });
        logger.debug('[OpenAIClient] chatCompletion: dropped params', {
-          dropParams: dropParams,
+          dropParams: this.options.dropParams,
          modelOptions,
        });
      }
@@ -1331,47 +1308,15 @@ ${convo}
      /** @type {(value: void | PromiseLike<void>) => void} */
      let streamResolve;

-      if (
-        (!this.isOmni || /^o1-(mini|preview)/i.test(modelOptions.model)) &&
-        modelOptions.reasoning_effort != null
-      ) {
-        delete modelOptions.reasoning_effort;
-        delete modelOptions.temperature;
-      }
-
-      let reasoningKey = 'reasoning_content';
-      if (this.useOpenRouter) {
-        modelOptions.include_reasoning = true;
-        reasoningKey = 'reasoning';
-      }
-      if (this.useOpenRouter && modelOptions.reasoning_effort != null) {
-        modelOptions.reasoning = {
-          effort: modelOptions.reasoning_effort,
-        };
-        delete modelOptions.reasoning_effort;
-      }
-
-      const handlers = createStreamEventHandlers(this.options.res);
-      this.streamHandler = new SplitStreamHandler({
-        reasoningKey,
-        accumulate: true,
-        runId: this.responseMessageId,
-        handlers,
-      });
-
-      intermediateReply = this.streamHandler.tokens;
-
      if (modelOptions.stream) {
        streamPromise = new Promise((resolve) => {
          streamResolve = resolve;
        });
-        /** @type {OpenAI.OpenAI.CompletionCreateParamsStreaming} */
-        const params = {
-          ...modelOptions,
-          stream: true,
-        };
-        const stream = await openai.chat.completions
-          .stream(params)
+        const stream = await openai.beta.chat.completions
+          .stream({
+            ...modelOptions,
+            stream: true,
+          })
          .on('abort', () => {
            /* Do nothing here */
          })
@@ -1389,44 +1334,20 @@ ${convo}
            }

            if (typeof finalMessage.content !== 'string' || finalMessage.content.trim() === '') {
-              finalChatCompletion.choices[0].message.content = this.streamHandler.tokens.join('');
+              finalChatCompletion.choices[0].message.content = intermediateReply.join('');
            }
          })
          .on('finalMessage', (message) => {
            if (message?.role !== 'assistant') {
-              stream.messages.push({
-                role: 'assistant',
-                content: this.streamHandler.tokens.join(''),
-              });
+              stream.messages.push({ role: 'assistant', content: intermediateReply.join('') });
              UnexpectedRoleError = true;
            }
          });

-        if (this.continued === true) {
-          const latestText = addSpaceIfNeeded(
-            this.currentMessages[this.currentMessages.length - 1]?.text ?? '',
-          );
-          this.streamHandler.handle({
-            choices: [
-              {
-                delta: {
-                  content: latestText,
-                },
-              },
-            ],
-          });
-        }
-
        for await (const chunk of stream) {
-          // Add finish_reason: null if missing in any choice
-          if (chunk.choices) {
-            chunk.choices.forEach((choice) => {
-              if (!('finish_reason' in choice)) {
-                choice.finish_reason = null;
-              }
-            });
-          }
-          this.streamHandler.handle(chunk);
+          const token = chunk.choices[0]?.delta?.content || '';
+          intermediateReply.push(token);
+          onProgress(token);
          if (abortController.signal.aborted) {
            stream.controller.abort();
            break;
@@ -1454,11 +1375,6 @@ ${convo}
          });
      }

-      if (openai.abortHandler && abortController.signal) {
-        abortController.signal.removeEventListener('abort', openai.abortHandler);
-        openai.abortHandler = undefined;
-      }
-
      if (!chatCompletion && UnexpectedRoleError) {
        throw new Error(
          'OpenAI error: Invalid final message: OpenAI expects final message to include role=assistant',
@@ -1474,7 +1390,7 @@ ${convo}

      if (!Array.isArray(choices) || choices.length === 0) {
        logger.warn('[OpenAIClient] Chat completion response has no choices');
-        return this.streamHandler.tokens.join('');
+        return intermediateReply.join('');
      }

      const { message, finish_reason } = choices[0] ?? {};
@@ -1484,11 +1400,11 @@ ${convo}

      if (!message) {
        logger.warn('[OpenAIClient] Message is undefined in chatCompletion response');
-        return this.streamHandler.tokens.join('');
+        return intermediateReply.join('');
      }

      if (typeof message.content !== 'string' || message.content.trim() === '') {
-        const reply = this.streamHandler.tokens.join('');
+        const reply = intermediateReply.join('');
        logger.debug(
          '[OpenAIClient] chatCompletion: using intermediateReply due to empty message.content',
          { intermediateReply: reply },
@@ -1496,27 +1412,13 @@ ${convo}
        return reply;
      }

-      if (
-        this.streamHandler.reasoningTokens.length > 0 &&
-        this.options.context !== 'title' &&
-        !message.content.startsWith('<think>')
-      ) {
-        return this.getStreamText();
-      } else if (
-        this.streamHandler.reasoningTokens.length > 0 &&
-        this.options.context !== 'title' &&
-        message.content.startsWith('<think>')
-      ) {
-        return this.getStreamText();
-      }
-
      return message.content;
    } catch (err) {
      if (
        err?.message?.includes('abort') ||
        (err instanceof OpenAI.APIError && err?.message?.includes('abort'))
      ) {
-        return this.getStreamText(intermediateReply);
+        return intermediateReply.join('');
      }
      if (
        err?.message?.includes(
@@ -1531,18 +1433,10 @@ ${convo}
        (err instanceof OpenAI.OpenAIError && err?.message?.includes('missing finish_reason'))
      ) {
        logger.error('[OpenAIClient] Known OpenAI error:', err);
-        if (this.streamHandler && this.streamHandler.reasoningTokens.length) {
-          return this.getStreamText();
-        } else if (intermediateReply.length > 0) {
-          return this.getStreamText(intermediateReply);
-        } else {
-          throw err;
-        }
+        return intermediateReply.join('');
      } else if (err instanceof OpenAI.APIError) {
-        if (this.streamHandler && this.streamHandler.reasoningTokens.length) {
-          return this.getStreamText();
-        } else if (intermediateReply.length > 0) {
-          return this.getStreamText(intermediateReply);
+        if (intermediateReply.length > 0) {
+          return intermediateReply.join('');
        } else {
          throw err;
        }
--- a/api/app/clients/PluginsClient.js
+++ b/api/app/clients/PluginsClient.js
@@ -0,0 +1,549 @@
+const OpenAIClient = require('./OpenAIClient');
+const { CacheKeys, Time } = require('librechat-data-provider');
+const { CallbackManager } = require('@langchain/core/callbacks/manager');
+const { BufferMemory, ChatMessageHistory } = require('langchain/memory');
+const { addImages, buildErrorInput, buildPromptPrefix } = require('./output_parsers');
+const { initializeCustomAgent, initializeFunctionsAgent } = require('./agents');
+const { processFileURL } = require('~/server/services/Files/process');
+const { EModelEndpoint } = require('librechat-data-provider');
+const { formatLangChainMessages } = require('./prompts');
+const checkBalance = require('~/models/checkBalance');
+const { isEnabled } = require('~/server/utils');
+const { extractBaseURL } = require('~/utils');
+const { loadTools } = require('./tools/util');
+const { getLogStores } = require('~/cache');
+const { logger } = require('~/config');
+
+class PluginsClient extends OpenAIClient {
+  constructor(apiKey, options = {}) {
+    super(apiKey, options);
+    this.sender = options.sender ?? 'Assistant';
+    this.tools = [];
+    this.actions = [];
+    this.setOptions(options);
+    this.openAIApiKey = this.apiKey;
+    this.executor = null;
+  }
+
+  setOptions(options) {
+    this.agentOptions = { ...options.agentOptions };
+    this.functionsAgent = this.agentOptions?.agent === 'functions';
+    this.agentIsGpt3 = this.agentOptions?.model?.includes('gpt-3');
+
+    super.setOptions(options);
+
+    this.isGpt3 = this.modelOptions?.model?.includes('gpt-3');
+
+    if (this.options.reverseProxyUrl) {
+      this.langchainProxy = extractBaseURL(this.options.reverseProxyUrl);
+    }
+  }
+
+  getSaveOptions() {
+    return {
+      artifacts: this.options.artifacts,
+      chatGptLabel: this.options.chatGptLabel,
+      promptPrefix: this.options.promptPrefix,
+      tools: this.options.tools,
+      ...this.modelOptions,
+      agentOptions: this.agentOptions,
+      iconURL: this.options.iconURL,
+      greeting: this.options.greeting,
+      spec: this.options.spec,
+    };
+  }
+
+  saveLatestAction(action) {
+    this.actions.push(action);
+  }
+
+  getFunctionModelName(input) {
+    if (/-(?!0314)\d{4}/.test(input)) {
+      return input;
+    } else if (input.includes('gpt-3.5-turbo')) {
+      return 'gpt-3.5-turbo';
+    } else if (input.includes('gpt-4')) {
+      return 'gpt-4';
+    } else {
+      return 'gpt-3.5-turbo';
+    }
+  }
+
+  getBuildMessagesOptions(opts) {
+    return {
+      isChatCompletion: true,
+      promptPrefix: opts.promptPrefix,
+      abortController: opts.abortController,
+    };
+  }
+
+  async initialize({ user, message, onAgentAction, onChainEnd, signal }) {
+    const modelOptions = {
+      modelName: this.agentOptions.model,
+      temperature: this.agentOptions.temperature,
+    };
+
+    const model = this.initializeLLM({
+      ...modelOptions,
+      context: 'plugins',
+      initialMessageCount: this.currentMessages.length + 1,
+    });
+
+    logger.debug(
+      `[PluginsClient] Agent Model: ${model.modelName} | Temp: ${model.temperature} | Functions: ${this.functionsAgent}`,
+    );
+
+    // Map Messages to Langchain format
+    const pastMessages = formatLangChainMessages(this.currentMessages.slice(0, -1), {
+      userName: this.options?.name,
+    });
+    logger.debug('[PluginsClient] pastMessages: ' + pastMessages.length);
+
+    // TODO: use readOnly memory, TokenBufferMemory? (both unavailable in LangChainJS)
+    const memory = new BufferMemory({
+      llm: model,
+      chatHistory: new ChatMessageHistory(pastMessages),
+    });
+
+    this.tools = await loadTools({
+      user,
+      model,
+      tools: this.options.tools,
+      functions: this.functionsAgent,
+      options: {
+        memory,
+        signal: this.abortController.signal,
+        openAIApiKey: this.openAIApiKey,
+        conversationId: this.conversationId,
+        fileStrategy: this.options.req.app.locals.fileStrategy,
+        processFileURL,
+        message,
+      },
+    });
+
+    if (this.tools.length === 0) {
+      return;
+    }
+
+    logger.debug('[PluginsClient] Requested Tools', this.options.tools);
+    logger.debug(
+      '[PluginsClient] Loaded Tools',
+      this.tools.map((tool) => tool.name),
+    );
+
+    const handleAction = (action, runId, callback = null) => {
+      this.saveLatestAction(action);
+
+      logger.debug('[PluginsClient] Latest Agent Action ', this.actions[this.actions.length - 1]);
+
+      if (typeof callback === 'function') {
+        callback(action, runId);
+      }
+    };
+
+    // initialize agent
+    const initializer = this.functionsAgent ? initializeFunctionsAgent : initializeCustomAgent;
+
+    let customInstructions = (this.options.promptPrefix ?? '').trim();
+    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
+      customInstructions = `${customInstructions ?? ''}\n${this.options.artifactsPrompt}`.trim();
+    }
+
+    this.executor = await initializer({
+      model,
+      signal,
+      pastMessages,
+      tools: this.tools,
+      customInstructions,
+      verbose: this.options.debug,
+      returnIntermediateSteps: true,
+      customName: this.options.chatGptLabel,
+      currentDateString: this.currentDateString,
+      callbackManager: CallbackManager.fromHandlers({
+        async handleAgentAction(action, runId) {
+          handleAction(action, runId, onAgentAction);
+        },
+        async handleChainEnd(action) {
+          if (typeof onChainEnd === 'function') {
+            onChainEnd(action);
+          }
+        },
+      }),
+    });
+
+    logger.debug('[PluginsClient] Loaded agent.');
+  }
+
+  async executorCall(message, { signal, stream, onToolStart, onToolEnd }) {
+    let errorMessage = '';
+    const maxAttempts = 1;
+
+    for (let attempts = 1; attempts <= maxAttempts; attempts++) {
+      const errorInput = buildErrorInput({
+        message,
+        errorMessage,
+        actions: this.actions,
+        functionsAgent: this.functionsAgent,
+      });
+      const input = attempts > 1 ? errorInput : message;
+
+      logger.debug(`[PluginsClient] Attempt ${attempts} of ${maxAttempts}`);
+
+      if (errorMessage.length > 0) {
+        logger.debug('[PluginsClient] Caught error, input: ' + JSON.stringify(input));
+      }
+
+      try {
+        this.result = await this.executor.call({ input, signal }, [
+          {
+            async handleToolStart(...args) {
+              await onToolStart(...args);
+            },
+            async handleToolEnd(...args) {
+              await onToolEnd(...args);
+            },
+            async handleLLMEnd(output) {
+              const { generations } = output;
+              const { text } = generations[0][0];
+              if (text && typeof stream === 'function') {
+                await stream(text);
+              }
+            },
+          },
+        ]);
+        break; // Exit the loop if the function call is successful
+      } catch (err) {
+        logger.error('[PluginsClient] executorCall error:', err);
+        if (attempts === maxAttempts) {
+          const { run } = this.runManager.getRunByConversationId(this.conversationId);
+          const defaultOutput = `Encountered an error while attempting to respond: ${err.message}`;
+          this.result.output = run && run.error ? run.error : defaultOutput;
+          this.result.errorMessage = run && run.error ? run.error : err.message;
+          this.result.intermediateSteps = this.actions;
+          break;
+        }
+      }
+    }
+  }
+
+  /**
+   *
+   * @param {TMessage} responseMessage
+   * @param {Partial<TMessage>} saveOptions
+   * @param {string} user
+   * @returns
+   */
+  async handleResponseMessage(responseMessage, saveOptions, user) {
+    const { output, errorMessage, ...result } = this.result;
+    logger.debug('[PluginsClient][handleResponseMessage] Output:', {
+      output,
+      errorMessage,
+      ...result,
+    });
+    const { error } = responseMessage;
+    if (!error) {
+      responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
+      responseMessage.completionTokens = this.getTokenCount(responseMessage.text);
+    }
+
+    // Record usage only when completion is skipped as it is already recorded in the agent phase.
+    if (!this.agentOptions.skipCompletion && !error) {
+      await this.recordTokenUsage(responseMessage);
+    }
+
+    this.responsePromise = this.saveMessageToDatabase(responseMessage, saveOptions, user);
+    const messageCache = getLogStores(CacheKeys.MESSAGES);
+    messageCache.set(
+      responseMessage.messageId,
+      {
+        text: responseMessage.text,
+        complete: true,
+      },
+      Time.FIVE_MINUTES,
+    );
+    delete responseMessage.tokenCount;
+    return { ...responseMessage, ...result };
+  }
+
+  async sendMessage(message, opts = {}) {
+    /** @type {{ filteredTools: string[], includedTools: string[] }} */
+    const { filteredTools = [], includedTools = [] } = this.options.req.app.locals;
+
+    if (includedTools.length > 0) {
+      const tools = this.options.tools.filter((plugin) => includedTools.includes(plugin));
+      this.options.tools = tools;
+    } else {
+      const tools = this.options.tools.filter((plugin) => !filteredTools.includes(plugin));
+      this.options.tools = tools;
+    }
+
+    // If a message is edited, no tools can be used.
+    const completionMode = this.options.tools.length === 0 || opts.isEdited;
+    if (completionMode) {
+      this.setOptions(opts);
+      return super.sendMessage(message, opts);
+    }
+
+    logger.debug('[PluginsClient] sendMessage', { userMessageText: message, opts });
+    const {
+      user,
+      isEdited,
+      conversationId,
+      responseMessageId,
+      saveOptions,
+      userMessage,
+      onAgentAction,
+      onChainEnd,
+      onToolStart,
+      onToolEnd,
+    } = await this.handleStartMethods(message, opts);
+
+    if (opts.progressCallback) {
+      opts.onProgress = opts.progressCallback.call(null, {
+        ...(opts.progressOptions ?? {}),
+        parentMessageId: userMessage.messageId,
+        messageId: responseMessageId,
+      });
+    }
+
+    this.currentMessages.push(userMessage);
+
+    let {
+      prompt: payload,
+      tokenCountMap,
+      promptTokens,
+    } = await this.buildMessages(
+      this.currentMessages,
+      userMessage.messageId,
+      this.getBuildMessagesOptions({
+        promptPrefix: null,
+        abortController: this.abortController,
+      }),
+    );
+
+    if (tokenCountMap) {
+      logger.debug('[PluginsClient] tokenCountMap', { tokenCountMap });
+      if (tokenCountMap[userMessage.messageId]) {
+        userMessage.tokenCount = tokenCountMap[userMessage.messageId];
+        logger.debug('[PluginsClient] userMessage.tokenCount', userMessage.tokenCount);
+      }
+      this.handleTokenCountMap(tokenCountMap);
+    }
+
+    this.result = {};
+    if (payload) {
+      this.currentMessages = payload;
+    }
+
+    if (!this.skipSaveUserMessage) {
+      this.userMessagePromise = this.saveMessageToDatabase(userMessage, saveOptions, user);
+      if (typeof opts?.getReqData === 'function') {
+        opts.getReqData({
+          userMessagePromise: this.userMessagePromise,
+        });
+      }
+    }
+
+    if (isEnabled(process.env.CHECK_BALANCE)) {
+      await checkBalance({
+        req: this.options.req,
+        res: this.options.res,
+        txData: {
+          user: this.user,
+          tokenType: 'prompt',
+          amount: promptTokens,
+          debug: this.options.debug,
+          model: this.modelOptions.model,
+          endpoint: EModelEndpoint.openAI,
+        },
+      });
+    }
+
+    const responseMessage = {
+      endpoint: EModelEndpoint.gptPlugins,
+      iconURL: this.options.iconURL,
+      messageId: responseMessageId,
+      conversationId,
+      parentMessageId: userMessage.messageId,
+      isCreatedByUser: false,
+      isEdited,
+      model: this.modelOptions.model,
+      sender: this.sender,
+      promptTokens,
+    };
+
+    await this.initialize({
+      user,
+      message,
+      onAgentAction,
+      onChainEnd,
+      signal: this.abortController.signal,
+      onProgress: opts.onProgress,
+    });
+
+    // const stream = async (text) => {
+    //   await this.generateTextStream.call(this, text, opts.onProgress, { delay: 1 });
+    // };
+    await this.executorCall(message, {
+      signal: this.abortController.signal,
+      // stream,
+      onToolStart,
+      onToolEnd,
+    });
+
+    // If message was aborted mid-generation
+    if (this.result?.errorMessage?.length > 0 && this.result?.errorMessage?.includes('cancel')) {
+      responseMessage.text = 'Cancelled.';
+      return await this.handleResponseMessage(responseMessage, saveOptions, user);
+    }
+
+    // If error occurred during generation (likely token_balance)
+    if (this.result?.errorMessage?.length > 0) {
+      responseMessage.error = true;
+      responseMessage.text = this.result.output;
+      return await this.handleResponseMessage(responseMessage, saveOptions, user);
+    }
+
+    if (this.agentOptions.skipCompletion && this.result.output && this.functionsAgent) {
+      const partialText = opts.getPartialText();
+      const trimmedPartial = opts.getPartialText().replaceAll(':::plugin:::\n', '');
+      responseMessage.text =
+        trimmedPartial.length === 0 ? `${partialText}${this.result.output}` : partialText;
+      addImages(this.result.intermediateSteps, responseMessage);
+      await this.generateTextStream(this.result.output, opts.onProgress, { delay: 5 });
+      return await this.handleResponseMessage(responseMessage, saveOptions, user);
+    }
+
+    if (this.agentOptions.skipCompletion && this.result.output) {
+      responseMessage.text = this.result.output;
+      addImages(this.result.intermediateSteps, responseMessage);
+      await this.generateTextStream(this.result.output, opts.onProgress, { delay: 5 });
+      return await this.handleResponseMessage(responseMessage, saveOptions, user);
+    }
+
+    logger.debug('[PluginsClient] Completion phase: this.result', this.result);
+
+    const promptPrefix = buildPromptPrefix({
+      result: this.result,
+      message,
+      functionsAgent: this.functionsAgent,
+    });
+
+    logger.debug('[PluginsClient]', { promptPrefix });
+
+    payload = await this.buildCompletionPrompt({
+      messages: this.currentMessages,
+      promptPrefix,
+    });
+
+    logger.debug('[PluginsClient] buildCompletionPrompt Payload', payload);
+    responseMessage.text = await this.sendCompletion(payload, opts);
+    return await this.handleResponseMessage(responseMessage, saveOptions, user);
+  }
+
+  async buildCompletionPrompt({ messages, promptPrefix: _promptPrefix }) {
+    logger.debug('[PluginsClient] buildCompletionPrompt messages', messages);
+
+    const orderedMessages = messages;
+    let promptPrefix = _promptPrefix.trim();
+    // If the prompt prefix doesn't end with the end token, add it.
+    if (!promptPrefix.endsWith(`${this.endToken}`)) {
+      promptPrefix = `${promptPrefix.trim()}${this.endToken}\n\n`;
+    }
+    promptPrefix = `${this.startToken}Instructions:\n${promptPrefix}`;
+    const promptSuffix = `${this.startToken}${this.chatGptLabel ?? 'Assistant'}:\n`;
+
+    const instructionsPayload = {
+      role: 'system',
+      content: promptPrefix,
+    };
+
+    const messagePayload = {
+      role: 'system',
+      content: promptSuffix,
+    };
+
+    if (this.isGpt3) {
+      instructionsPayload.role = 'user';
+      messagePayload.role = 'user';
+      instructionsPayload.content += `\n${promptSuffix}`;
+    }
+
+    // testing if this works with browser endpoint
+    if (!this.isGpt3 && this.options.reverseProxyUrl) {
+      instructionsPayload.role = 'user';
+    }
+
+    let currentTokenCount =
+      this.getTokenCountForMessage(instructionsPayload) +
+      this.getTokenCountForMessage(messagePayload);
+
+    let promptBody = '';
+    const maxTokenCount = this.maxPromptTokens;
+    // Iterate backwards through the messages, adding them to the prompt until we reach the max token count.
+    // Do this within a recursive async function so that it doesn't block the event loop for too long.
+    const buildPromptBody = async () => {
+      if (currentTokenCount < maxTokenCount && orderedMessages.length > 0) {
+        const message = orderedMessages.pop();
+        const isCreatedByUser = message.isCreatedByUser || message.role?.toLowerCase() === 'user';
+        const roleLabel = isCreatedByUser ? this.userLabel : this.chatGptLabel;
+        let messageString = `${this.startToken}${roleLabel}:\n${
+          message.text ?? message.content ?? ''
+        }${this.endToken}\n`;
+        let newPromptBody = `${messageString}${promptBody}`;
+
+        const tokenCountForMessage = this.getTokenCount(messageString);
+        const newTokenCount = currentTokenCount + tokenCountForMessage;
+        if (newTokenCount > maxTokenCount) {
+          if (promptBody) {
+            // This message would put us over the token limit, so don't add it.
+            return false;
+          }
+          // This is the first message, so we can't add it. Just throw an error.
+          throw new Error(
+            `Prompt is too long. Max token count is ${maxTokenCount}, but prompt is ${newTokenCount} tokens long.`,
+          );
+        }
+        promptBody = newPromptBody;
+        currentTokenCount = newTokenCount;
+        // wait for next tick to avoid blocking the event loop
+        await new Promise((resolve) => setTimeout(resolve, 0));
+        return buildPromptBody();
+      }
+      return true;
+    };
+
+    await buildPromptBody();
+    const prompt = promptBody;
+    messagePayload.content = prompt;
+    // Add 2 tokens for metadata after all messages have been counted.
+    currentTokenCount += 2;
+
+    if (this.isGpt3 && messagePayload.content.length > 0) {
+      const context = 'Chat History:\n';
+      messagePayload.content = `${context}${prompt}`;
+      currentTokenCount += this.getTokenCount(context);
+    }
+
+    // Use up to `this.maxContextTokens` tokens (prompt + response), but try to leave `this.maxTokens` tokens for the response.
+    this.modelOptions.max_tokens = Math.min(
+      this.maxContextTokens - currentTokenCount,
+      this.maxResponseTokens,
+    );
+
+    if (this.isGpt3) {
+      messagePayload.content += promptSuffix;
+      return [instructionsPayload, messagePayload];
+    }
+
+    const result = [messagePayload, instructionsPayload];
+
+    if (this.functionsAgent && !this.isGpt3) {
+      result[1].content = `${result[1].content}\n${this.startToken}${this.chatGptLabel}:\nSure thing! Here is the output you requested:\n`;
+    }
+
+    return result.filter((message) => message.content.length > 0);
+  }
+}
+
+module.exports = PluginsClient;
--- a/api/app/clients/callbacks/createStartHandler.js
+++ b/api/app/clients/callbacks/createStartHandler.js
@@ -0,0 +1,95 @@
+const { promptTokensEstimate } = require('openai-chat-tokens');
+const { EModelEndpoint, supportsBalanceCheck } = require('librechat-data-provider');
+const { formatFromLangChain } = require('~/app/clients/prompts');
+const checkBalance = require('~/models/checkBalance');
+const { isEnabled } = require('~/server/utils');
+const { logger } = require('~/config');
+
+const createStartHandler = ({
+  context,
+  conversationId,
+  tokenBuffer = 0,
+  initialMessageCount,
+  manager,
+}) => {
+  return async (_llm, _messages, runId, parentRunId, extraParams) => {
+    const { invocation_params } = extraParams;
+    const { model, functions, function_call } = invocation_params;
+    const messages = _messages[0].map(formatFromLangChain);
+
+    logger.debug(`[createStartHandler] handleChatModelStart: ${context}`, {
+      model,
+      function_call,
+    });
+
+    if (context !== 'title') {
+      logger.debug(`[createStartHandler] handleChatModelStart: ${context}`, {
+        functions,
+      });
+    }
+
+    const payload = { messages };
+    let prelimPromptTokens = 1;
+
+    if (functions) {
+      payload.functions = functions;
+      prelimPromptTokens += 2;
+    }
+
+    if (function_call) {
+      payload.function_call = function_call;
+      prelimPromptTokens -= 5;
+    }
+
+    prelimPromptTokens += promptTokensEstimate(payload);
+    logger.debug('[createStartHandler]', {
+      prelimPromptTokens,
+      tokenBuffer,
+    });
+    prelimPromptTokens += tokenBuffer;
+
+    try {
+      // TODO: if plugins extends to non-OpenAI models, this will need to be updated
+      if (isEnabled(process.env.CHECK_BALANCE) && supportsBalanceCheck[EModelEndpoint.openAI]) {
+        const generations =
+          initialMessageCount && messages.length > initialMessageCount
+            ? messages.slice(initialMessageCount)
+            : null;
+        await checkBalance({
+          req: manager.req,
+          res: manager.res,
+          txData: {
+            user: manager.user,
+            tokenType: 'prompt',
+            amount: prelimPromptTokens,
+            debug: manager.debug,
+            generations,
+            model,
+            endpoint: EModelEndpoint.openAI,
+          },
+        });
+      }
+    } catch (err) {
+      logger.error(`[createStartHandler][${context}] checkBalance error`, err);
+      manager.abortController.abort();
+      if (context === 'summary' || context === 'plugins') {
+        manager.addRun(runId, { conversationId, error: err.message });
+        throw new Error(err);
+      }
+      return;
+    }
+
+    manager.addRun(runId, {
+      model,
+      messages,
+      functions,
+      function_call,
+      runId,
+      parentRunId,
+      conversationId,
+      prelimPromptTokens,
+    });
+  };
+};
+
+module.exports = createStartHandler;
--- a/api/app/clients/callbacks/index.js
+++ b/api/app/clients/callbacks/index.js
@@ -0,0 +1,5 @@
+const createStartHandler = require('./createStartHandler');
+
+module.exports = {
+  createStartHandler,
+};
--- a/api/app/clients/index.js
+++ b/api/app/clients/index.js
@@ -1,11 +1,15 @@
+const ChatGPTClient = require('./ChatGPTClient');
 const OpenAIClient = require('./OpenAIClient');
+const PluginsClient = require('./PluginsClient');
 const GoogleClient = require('./GoogleClient');
 const TextStream = require('./TextStream');
 const AnthropicClient = require('./AnthropicClient');
 const toolUtils = require('./tools/util');

 module.exports = {
+  ChatGPTClient,
  OpenAIClient,
+  PluginsClient,
  GoogleClient,
  TextStream,
  AnthropicClient,
--- a/api/app/clients/llm/RunManager.js
+++ b/api/app/clients/llm/RunManager.js
@@ -0,0 +1,105 @@
+const { createStartHandler } = require('~/app/clients/callbacks');
+const { spendTokens } = require('~/models/spendTokens');
+const { logger } = require('~/config');
+
+class RunManager {
+  constructor(fields) {
+    const { req, res, abortController, debug } = fields;
+    this.abortController = abortController;
+    this.user = req.user.id;
+    this.req = req;
+    this.res = res;
+    this.debug = debug;
+    this.runs = new Map();
+    this.convos = new Map();
+  }
+
+  addRun(runId, runData) {
+    if (!this.runs.has(runId)) {
+      this.runs.set(runId, runData);
+      if (runData.conversationId) {
+        this.convos.set(runData.conversationId, runId);
+      }
+      return runData;
+    } else {
+      const existingData = this.runs.get(runId);
+      const update = { ...existingData, ...runData };
+      this.runs.set(runId, update);
+      if (update.conversationId) {
+        this.convos.set(update.conversationId, runId);
+      }
+      return update;
+    }
+  }
+
+  removeRun(runId) {
+    if (this.runs.has(runId)) {
+      this.runs.delete(runId);
+    } else {
+      logger.error(`[api/app/clients/llm/RunManager] Run with ID ${runId} does not exist.`);
+    }
+  }
+
+  getAllRuns() {
+    return Array.from(this.runs.values());
+  }
+
+  getRunById(runId) {
+    return this.runs.get(runId);
+  }
+
+  getRunByConversationId(conversationId) {
+    const runId = this.convos.get(conversationId);
+    return { run: this.runs.get(runId), runId };
+  }
+
+  createCallbacks(metadata) {
+    return [
+      {
+        handleChatModelStart: createStartHandler({ ...metadata, manager: this }),
+        handleLLMEnd: async (output, runId, _parentRunId) => {
+          const { llmOutput, ..._output } = output;
+          logger.debug(`[RunManager] handleLLMEnd: ${JSON.stringify(metadata)}`, {
+            runId,
+            _parentRunId,
+            llmOutput,
+          });
+
+          if (metadata.context !== 'title') {
+            logger.debug('[RunManager] handleLLMEnd:', {
+              output: _output,
+            });
+          }
+
+          const { tokenUsage } = output.llmOutput;
+          const run = this.getRunById(runId);
+          this.removeRun(runId);
+
+          const txData = {
+            user: this.user,
+            model: run?.model ?? 'gpt-3.5-turbo',
+            ...metadata,
+          };
+
+          await spendTokens(txData, tokenUsage);
+        },
+        handleLLMError: async (err) => {
+          logger.error(`[RunManager] handleLLMError: ${JSON.stringify(metadata)}`, err);
+          if (metadata.context === 'title') {
+            return;
+          } else if (metadata.context === 'plugins') {
+            throw new Error(err);
+          }
+          const { conversationId } = metadata;
+          const { run } = this.getRunByConversationId(conversationId);
+          if (run && run.error) {
+            const { error } = run;
+            throw new Error(error);
+          }
+        },
+      },
+    ];
+  }
+}
+
+module.exports = RunManager;
--- a/api/app/clients/llm/createLLM.js
+++ b/api/app/clients/llm/createLLM.js
@@ -1,5 +1,6 @@
 const { ChatOpenAI } = require('@langchain/openai');
-const { isEnabled, sanitizeModelName, constructAzureURL } = require('@librechat/api');
+const { sanitizeModelName, constructAzureURL } = require('~/utils');
+const { isEnabled } = require('~/server/utils');

 /**
 * Creates a new instance of a language model (LLM) for chat interactions.
@@ -16,7 +17,7 @@ const { isEnabled, sanitizeModelName, constructAzureURL } = require('@librechat/
 *
 * @example
 * const llm = createLLM({
- *   modelOptions: { modelName: 'gpt-4o-mini', temperature: 0.2 },
+ *   modelOptions: { modelName: 'gpt-3.5-turbo', temperature: 0.2 },
 *   configOptions: { basePath: 'https://example.api/path' },
 *   callbacks: { onMessage: handleMessage },
 *   openAIApiKey: 'your-api-key'
@@ -33,7 +34,6 @@ function createLLM({
  let credentials = { openAIApiKey };
  let configuration = {
    apiKey: openAIApiKey,
-    ...(configOptions.basePath && { baseURL: configOptions.basePath }),
  };

  /**  @type {AzureOptions} */
--- a/api/app/clients/llm/index.js
+++ b/api/app/clients/llm/index.js
@@ -1,7 +1,9 @@
 const createLLM = require('./createLLM');
+const RunManager = require('./RunManager');
 const createCoherePayload = require('./createCoherePayload');

 module.exports = {
  createLLM,
+  RunManager,
  createCoherePayload,
 };
--- a/api/app/clients/memory/summaryBuffer.demo.js
+++ b/api/app/clients/memory/summaryBuffer.demo.js
@@ -3,7 +3,7 @@ const { ChatOpenAI } = require('@langchain/openai');
 const { getBufferString, ConversationSummaryBufferMemory } = require('langchain/memory');

 const chatPromptMemory = new ConversationSummaryBufferMemory({
-  llm: new ChatOpenAI({ modelName: 'gpt-4o-mini', temperature: 0 }),
+  llm: new ChatOpenAI({ modelName: 'gpt-3.5-turbo', temperature: 0 }),
  maxTokenLimit: 10,
  returnMessages: true,
 });
--- a/api/app/clients/prompts/addCacheControl.js
+++ b/api/app/clients/prompts/addCacheControl.js
@@ -1,7 +1,7 @@
 /**
 * Anthropic API: Adds cache control to the appropriate user messages in the payload.
- * @param {Array<AnthropicMessage | BaseMessage>} messages - The array of message objects.
- * @returns {Array<AnthropicMessage | BaseMessage>} - The updated array of message objects with cache control added.
+ * @param {Array<AnthropicMessage>} messages - The array of message objects.
+ * @returns {Array<AnthropicMessage>} - The updated array of message objects with cache control added.
 */
 function addCacheControl(messages) {
  if (!Array.isArray(messages) || messages.length < 2) {
@@ -13,9 +13,7 @@ function addCacheControl(messages) {

  for (let i = updatedMessages.length - 1; i >= 0 && userMessagesModified < 2; i--) {
    const message = updatedMessages[i];
-    if (message.getType != null && message.getType() !== 'human') {
-      continue;
-    } else if (message.getType == null && message.role !== 'user') {
+    if (message.role !== 'user') {
      continue;
    }

--- a/api/app/clients/prompts/createContextHandlers.js
+++ b/api/app/clients/prompts/createContextHandlers.js
@@ -1,6 +1,6 @@
 const axios = require('axios');
-const { logger } = require('@librechat/data-schemas');
-const { isEnabled, generateShortLivedToken } = require('@librechat/api');
+const { isEnabled } = require('~/server/utils');
+const { logger } = require('~/config');

 const footer = `Use the context as your learned knowledge to better answer the user.

@@ -18,7 +18,7 @@ function createContextHandlers(req, userMessageContent) {
  const queryPromises = [];
  const processedFiles = [];
  const processedIds = new Set();
-  const jwtToken = generateShortLivedToken(req.user.id);
+  const jwtToken = req.headers.authorization.split(' ')[1];
  const useFullContext = isEnabled(process.env.RAG_USE_FULL_CONTEXT);

  const query = async (file) => {
@@ -96,35 +96,35 @@ function createContextHandlers(req, userMessageContent) {
        resolvedQueries.length === 0
          ? '\n\tThe semantic search did not return any results.'
          : resolvedQueries
-              .map((queryResult, index) => {
-                const file = processedFiles[index];
-                let contextItems = queryResult.data;
+            .map((queryResult, index) => {
+              const file = processedFiles[index];
+              let contextItems = queryResult.data;

-                const generateContext = (currentContext) =>
-                  `
+              const generateContext = (currentContext) =>
+                `
          <file>
            <filename>${file.filename}</filename>
            <context>${currentContext}
            </context>
          </file>`;

-                if (useFullContext) {
-                  return generateContext(`\n${contextItems}`);
-                }
+              if (useFullContext) {
+                return generateContext(`\n${contextItems}`);
+              }

-                contextItems = queryResult.data
-                  .map((item) => {
-                    const pageContent = item[0].page_content;
-                    return `
+              contextItems = queryResult.data
+                .map((item) => {
+                  const pageContent = item[0].page_content;
+                  return `
            <contextItem>
              <![CDATA[${pageContent?.trim()}]]>
            </contextItem>`;
-                  })
-                  .join('');
+                })
+                .join('');

-                return generateContext(contextItems);
-              })
-              .join('');
+              return generateContext(contextItems);
+            })
+            .join('');

      if (useFullContext) {
        const prompt = `${header}
--- a/api/app/clients/prompts/formatAgentMessages.spec.js
+++ b/api/app/clients/prompts/formatAgentMessages.spec.js
@@ -282,80 +282,4 @@ describe('formatAgentMessages', () => {
    // Additional check to ensure the consecutive assistant messages were combined
    expect(result[1].content).toHaveLength(2);
  });
-
-  it('should skip THINK type content parts', () => {
-    const payload = [
-      {
-        role: 'assistant',
-        content: [
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Initial response' },
-          { type: ContentTypes.THINK, [ContentTypes.THINK]: 'Reasoning about the problem...' },
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Final answer' },
-        ],
-      },
-    ];
-
-    const result = formatAgentMessages(payload);
-
-    expect(result).toHaveLength(1);
-    expect(result[0]).toBeInstanceOf(AIMessage);
-    expect(result[0].content).toEqual('Initial response\nFinal answer');
-  });
-
-  it('should join TEXT content as string when THINK content type is present', () => {
-    const payload = [
-      {
-        role: 'assistant',
-        content: [
-          { type: ContentTypes.THINK, [ContentTypes.THINK]: 'Analyzing the problem...' },
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'First part of response' },
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Second part of response' },
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Final part of response' },
-        ],
-      },
-    ];
-
-    const result = formatAgentMessages(payload);
-
-    expect(result).toHaveLength(1);
-    expect(result[0]).toBeInstanceOf(AIMessage);
-    expect(typeof result[0].content).toBe('string');
-    expect(result[0].content).toBe(
-      'First part of response\nSecond part of response\nFinal part of response',
-    );
-    expect(result[0].content).not.toContain('Analyzing the problem...');
-  });
-
-  it('should exclude ERROR type content parts', () => {
-    const payload = [
-      {
-        role: 'assistant',
-        content: [
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Hello there' },
-          {
-            type: ContentTypes.ERROR,
-            [ContentTypes.ERROR]:
-              'An error occurred while processing the request: Something went wrong',
-          },
-          { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Final answer' },
-        ],
-      },
-    ];
-
-    const result = formatAgentMessages(payload);
-
-    expect(result).toHaveLength(1);
-    expect(result[0]).toBeInstanceOf(AIMessage);
-    expect(result[0].content).toEqual([
-      { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Hello there' },
-      { type: ContentTypes.TEXT, [ContentTypes.TEXT]: 'Final answer' },
-    ]);
-
-    // Make sure no error content exists in the result
-    const hasErrorContent = result[0].content.some(
-      (item) =>
-        item.type === ContentTypes.ERROR || JSON.stringify(item).includes('An error occurred'),
-    );
-    expect(hasErrorContent).toBe(false);
-  });
 });
--- a/api/app/clients/prompts/formatMessages.js
+++ b/api/app/clients/prompts/formatMessages.js
@@ -153,7 +153,6 @@ const formatAgentMessages = (payload) => {
    let currentContent = [];
    let lastAIMessage = null;

-    let hasReasoning = false;
    for (const part of message.content) {
      if (part.type === ContentTypes.TEXT && part.tool_call_ids) {
        /*
@@ -205,30 +204,14 @@ const formatAgentMessages = (payload) => {
          new ToolMessage({
            tool_call_id: tool_call.id,
            name: tool_call.name,
-            content: output || '',
+            content: output,
          }),
        );
-      } else if (part.type === ContentTypes.THINK) {
-        hasReasoning = true;
-        continue;
-      } else if (part.type === ContentTypes.ERROR || part.type === ContentTypes.AGENT_UPDATE) {
-        continue;
      } else {
        currentContent.push(part);
      }
    }

-    if (hasReasoning) {
-      currentContent = currentContent
-        .reduce((acc, curr) => {
-          if (curr.type === ContentTypes.TEXT) {
-            return `${acc}${curr[ContentTypes.TEXT]}\n`;
-          }
-          return acc;
-        }, '')
-        .trim();
-    }
-
    if (currentContent.length > 0) {
      messages.push(new AIMessage({ content: currentContent }));
    }
@@ -237,9 +220,41 @@ const formatAgentMessages = (payload) => {
  return messages;
 };

+/**
+ * Formats an array of messages for LangChain, making sure all content fields are strings
+ * @param {Array<(HumanMessage|AIMessage|SystemMessage|ToolMessage)>} payload - The array of messages to format.
+ * @returns {Array<(HumanMessage|AIMessage|SystemMessage|ToolMessage)>} - The array of formatted LangChain messages, including ToolMessages for tool calls.
+ */
+const formatContentStrings = (payload) => {
+  const messages = [];
+
+  for (const message of payload) {
+    if (typeof message.content === 'string') {
+      continue;
+    }
+
+    if (!Array.isArray(message.content)) {
+      continue;
+    }
+
+    // Reduce text types to a single string, ignore all other types
+    const content = message.content.reduce((acc, curr) => {
+      if (curr.type === ContentTypes.TEXT) {
+        return `${acc}${curr[ContentTypes.TEXT]}\n`;
+      }
+      return acc;
+    }, '');
+
+    message.content = content.trim();
+  }
+
+  return messages;
+};
+
 module.exports = {
  formatMessage,
  formatFromLangChain,
  formatAgentMessages,
+  formatContentStrings,
  formatLangChainMessages,
 };
--- a/api/app/clients/prompts/formatMessages.spec.js
+++ b/api/app/clients/prompts/formatMessages.spec.js
@@ -60,6 +60,7 @@ describe('formatMessage', () => {
        error: false,
        finish_reason: null,
        isCreatedByUser: true,
+        isEdited: false,
        model: null,
        parentMessageId: Constants.NO_PARENT,
        sender: 'User',
--- a/api/app/clients/prompts/index.js
+++ b/api/app/clients/prompts/index.js
@@ -4,7 +4,7 @@ const summaryPrompts = require('./summaryPrompts');
 const handleInputs = require('./handleInputs');
 const instructions = require('./instructions');
 const titlePrompts = require('./titlePrompts');
-const truncate = require('./truncate');
+const truncateText = require('./truncateText');
 const createVisionPrompt = require('./createVisionPrompt');
 const createContextHandlers = require('./createContextHandlers');

@@ -15,7 +15,7 @@ module.exports = {
  ...handleInputs,
  ...instructions,
  ...titlePrompts,
-  ...truncate,
+  ...truncateText,
  createVisionPrompt,
  createContextHandlers,
 };
--- a/api/app/clients/prompts/truncate.js
+++ b/api/app/clients/prompts/truncate.js
@@ -1,115 +0,0 @@
-const MAX_CHAR = 255;
-
-/**
- * Truncates a given text to a specified maximum length, appending ellipsis and a notification
- * if the original text exceeds the maximum length.
- *
- * @param {string} text - The text to be truncated.
- * @param {number} [maxLength=MAX_CHAR] - The maximum length of the text after truncation. Defaults to MAX_CHAR.
- * @returns {string} The truncated text if the original text length exceeds maxLength, otherwise returns the original text.
- */
-function truncateText(text, maxLength = MAX_CHAR) {
-  if (text.length > maxLength) {
-    return `${text.slice(0, maxLength)}... [text truncated for brevity]`;
-  }
-  return text;
-}
-
-/**
- * Truncates a given text to a specified maximum length by showing the first half and the last half of the text,
- * separated by ellipsis. This method ensures the output does not exceed the maximum length, including the addition
- * of ellipsis and notification if the original text exceeds the maximum length.
- *
- * @param {string} text - The text to be truncated.
- * @param {number} [maxLength=MAX_CHAR] - The maximum length of the output text after truncation. Defaults to MAX_CHAR.
- * @returns {string} The truncated text showing the first half and the last half, or the original text if it does not exceed maxLength.
- */
-function smartTruncateText(text, maxLength = MAX_CHAR) {
-  const ellipsis = '...';
-  const notification = ' [text truncated for brevity]';
-  const halfMaxLength = Math.floor((maxLength - ellipsis.length - notification.length) / 2);
-
-  if (text.length > maxLength) {
-    const startLastHalf = text.length - halfMaxLength;
-    return `${text.slice(0, halfMaxLength)}${ellipsis}${text.slice(startLastHalf)}${notification}`;
-  }
-
-  return text;
-}
-
-/**
- * @param {TMessage[]} _messages
- * @param {number} maxContextTokens
- * @param {function({role: string, content: TMessageContent[]}): number} getTokenCountForMessage
- *
- * @returns {{
- *  dbMessages: TMessage[],
- * editedIndices: number[]
- * }}
- */
-function truncateToolCallOutputs(_messages, maxContextTokens, getTokenCountForMessage) {
-  const THRESHOLD_PERCENTAGE = 0.5;
-  const targetTokenLimit = maxContextTokens * THRESHOLD_PERCENTAGE;
-
-  let currentTokenCount = 3;
-  const messages = [..._messages];
-  const processedMessages = [];
-  let currentIndex = messages.length;
-  const editedIndices = new Set();
-  while (messages.length > 0) {
-    currentIndex--;
-    const message = messages.pop();
-    currentTokenCount += message.tokenCount;
-    if (currentTokenCount < targetTokenLimit) {
-      processedMessages.push(message);
-      continue;
-    }
-
-    if (!message.content || !Array.isArray(message.content)) {
-      processedMessages.push(message);
-      continue;
-    }
-
-    const toolCallIndices = message.content
-      .map((item, index) => (item.type === 'tool_call' ? index : -1))
-      .filter((index) => index !== -1)
-      .reverse();
-
-    if (toolCallIndices.length === 0) {
-      processedMessages.push(message);
-      continue;
-    }
-
-    const newContent = [...message.content];
-
-    // Truncate all tool outputs since we're over threshold
-    for (const index of toolCallIndices) {
-      const toolCall = newContent[index].tool_call;
-      if (!toolCall || !toolCall.output) {
-        continue;
-      }
-
-      editedIndices.add(currentIndex);
-
-      newContent[index] = {
-        ...newContent[index],
-        tool_call: {
-          ...toolCall,
-          output: '[OUTPUT_OMITTED_FOR_BREVITY]',
-        },
-      };
-    }
-
-    const truncatedMessage = {
-      ...message,
-      content: newContent,
-      tokenCount: getTokenCountForMessage({ role: 'assistant', content: newContent }),
-    };
-
-    processedMessages.push(truncatedMessage);
-  }
-
-  return { dbMessages: processedMessages.reverse(), editedIndices: Array.from(editedIndices) };
-}
-
-module.exports = { truncateText, smartTruncateText, truncateToolCallOutputs };
--- a/api/app/clients/prompts/truncateText.js
+++ b/api/app/clients/prompts/truncateText.js
@@ -0,0 +1,40 @@
+const MAX_CHAR = 255;
+
+/**
+ * Truncates a given text to a specified maximum length, appending ellipsis and a notification
+ * if the original text exceeds the maximum length.
+ *
+ * @param {string} text - The text to be truncated.
+ * @param {number} [maxLength=MAX_CHAR] - The maximum length of the text after truncation. Defaults to MAX_CHAR.
+ * @returns {string} The truncated text if the original text length exceeds maxLength, otherwise returns the original text.
+ */
+function truncateText(text, maxLength = MAX_CHAR) {
+  if (text.length > maxLength) {
+    return `${text.slice(0, maxLength)}... [text truncated for brevity]`;
+  }
+  return text;
+}
+
+/**
+ * Truncates a given text to a specified maximum length by showing the first half and the last half of the text,
+ * separated by ellipsis. This method ensures the output does not exceed the maximum length, including the addition
+ * of ellipsis and notification if the original text exceeds the maximum length.
+ *
+ * @param {string} text - The text to be truncated.
+ * @param {number} [maxLength=MAX_CHAR] - The maximum length of the output text after truncation. Defaults to MAX_CHAR.
+ * @returns {string} The truncated text showing the first half and the last half, or the original text if it does not exceed maxLength.
+ */
+function smartTruncateText(text, maxLength = MAX_CHAR) {
+  const ellipsis = '...';
+  const notification = ' [text truncated for brevity]';
+  const halfMaxLength = Math.floor((maxLength - ellipsis.length - notification.length) / 2);
+
+  if (text.length > maxLength) {
+    const startLastHalf = text.length - halfMaxLength;
+    return `${text.slice(0, halfMaxLength)}${ellipsis}${text.slice(startLastHalf)}${notification}`;
+  }
+
+  return text;
+}
+
+module.exports = { truncateText, smartTruncateText };
--- a/api/app/clients/specs/AnthropicClient.test.js
+++ b/api/app/clients/specs/AnthropicClient.test.js
@@ -1,4 +1,3 @@
-const { SplitStreamHandler } = require('@librechat/agents');
 const { anthropicSettings } = require('librechat-data-provider');
 const AnthropicClient = require('~/app/clients/AnthropicClient');

@@ -15,7 +14,7 @@ describe('AnthropicClient', () => {
    {
      role: 'user',
      isCreatedByUser: true,
-      text: "What's up",
+      text: 'What\'s up',
      messageId: '3',
      parentMessageId: '2',
    },
@@ -170,7 +169,7 @@ describe('AnthropicClient', () => {
      client.options.modelLabel = 'Claude-2';
      const result = await client.buildMessages(messages, parentMessageId);
      const { prompt } = result;
-      expect(prompt).toContain("Human's name: John");
+      expect(prompt).toContain('Human\'s name: John');
      expect(prompt).toContain('You are Claude-2');
    });
  });
@@ -244,70 +243,6 @@ describe('AnthropicClient', () => {
      );
    });

-    describe('Claude 4 model headers', () => {
-      it('should add "prompt-caching" and "context-1m" beta headers for claude-sonnet-4 model', () => {
-        const client = new AnthropicClient('test-api-key');
-        const modelOptions = {
-          model: 'claude-sonnet-4-20250514',
-        };
-        client.setOptions({ modelOptions, promptCache: true });
-        const anthropicClient = client.getClient(modelOptions);
-        expect(anthropicClient._options.defaultHeaders).toBeDefined();
-        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
-        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
-          'prompt-caching-2024-07-31,context-1m-2025-08-07',
-        );
-      });
-
-      it('should add "prompt-caching" and "context-1m" beta headers for claude-sonnet-4 model formats', () => {
-        const client = new AnthropicClient('test-api-key');
-        const modelVariations = [
-          'claude-sonnet-4-20250514',
-          'claude-sonnet-4-latest',
-          'anthropic/claude-sonnet-4-20250514',
-        ];
-
-        modelVariations.forEach((model) => {
-          const modelOptions = { model };
-          client.setOptions({ modelOptions, promptCache: true });
-          const anthropicClient = client.getClient(modelOptions);
-          expect(anthropicClient._options.defaultHeaders).toBeDefined();
-          expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
-          expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
-            'prompt-caching-2024-07-31,context-1m-2025-08-07',
-          );
-        });
-      });
-
-      it('should add "prompt-caching" beta header for claude-opus-4 model', () => {
-        const client = new AnthropicClient('test-api-key');
-        const modelOptions = {
-          model: 'claude-opus-4-20250514',
-        };
-        client.setOptions({ modelOptions, promptCache: true });
-        const anthropicClient = client.getClient(modelOptions);
-        expect(anthropicClient._options.defaultHeaders).toBeDefined();
-        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
-        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
-          'prompt-caching-2024-07-31',
-        );
-      });
-
-      it('should add "prompt-caching" beta header for claude-4-opus model', () => {
-        const client = new AnthropicClient('test-api-key');
-        const modelOptions = {
-          model: 'claude-4-opus-20250514',
-        };
-        client.setOptions({ modelOptions, promptCache: true });
-        const anthropicClient = client.getClient(modelOptions);
-        expect(anthropicClient._options.defaultHeaders).toBeDefined();
-        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
-        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
-          'prompt-caching-2024-07-31',
-        );
-      });
-    });
-
    it('should not add beta header for claude-3-5-sonnet-latest model', () => {
      const client = new AnthropicClient('test-api-key');
      const modelOptions = {
@@ -315,7 +250,7 @@ describe('AnthropicClient', () => {
      };
      client.setOptions({ modelOptions, promptCache: true });
      const anthropicClient = client.getClient(modelOptions);
-      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
+      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
    });

    it('should not add beta header for other models', () => {
@@ -326,7 +261,7 @@ describe('AnthropicClient', () => {
        },
      });
      const anthropicClient = client.getClient();
-      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
+      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
    });
  });

@@ -470,574 +405,4 @@ describe('AnthropicClient', () => {
      expect(Number.isNaN(result)).toBe(false);
    });
  });
-
-  describe('maxOutputTokens handling for different models', () => {
-    it('should not cap maxOutputTokens for Claude 3.5 Sonnet models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 10;
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-5-sonnet',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-
-      // Test with decimal notation
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.5-sonnet',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-    });
-
-    it('should not cap maxOutputTokens for Claude 3.7 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 2;
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-7-sonnet',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-
-      // Test with decimal notation
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.7-sonnet',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-    });
-
-    it('should not cap maxOutputTokens for Claude 4 Sonnet models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 10; // 40,960 tokens
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-sonnet-4-20250514',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-    });
-
-    it('should not cap maxOutputTokens for Claude 4 Opus models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 6; // 24,576 tokens (under 32K limit)
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-opus-4-20250514',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
-    });
-
-    it('should cap maxOutputTokens for Claude 3.5 Haiku models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 2;
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-5-haiku',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(
-        anthropicSettings.legacy.maxOutputTokens.default,
-      );
-
-      // Test with decimal notation
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.5-haiku',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(
-        anthropicSettings.legacy.maxOutputTokens.default,
-      );
-    });
-
-    it('should cap maxOutputTokens for Claude 3 Haiku and Opus models', () => {
-      const client = new AnthropicClient('test-api-key');
-      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 2;
-
-      // Test haiku
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-haiku',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(
-        anthropicSettings.legacy.maxOutputTokens.default,
-      );
-
-      // Test opus
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-opus',
-          maxOutputTokens: highTokenValue,
-        },
-      });
-
-      expect(client.modelOptions.maxOutputTokens).toBe(
-        anthropicSettings.legacy.maxOutputTokens.default,
-      );
-    });
-  });
-
-  describe('topK/topP parameters for different models', () => {
-    beforeEach(() => {
-      // Mock the SplitStreamHandler
-      jest.spyOn(SplitStreamHandler.prototype, 'handle').mockImplementation(() => {});
-    });
-
-    afterEach(() => {
-      jest.restoreAllMocks();
-    });
-
-    it('should include top_k and top_p parameters for non-claude-3.7 models', async () => {
-      const client = new AnthropicClient('test-api-key');
-
-      // Create a mock async generator function
-      async function* mockAsyncGenerator() {
-        yield { type: 'message_start', message: { usage: {} } };
-        yield { delta: { text: 'Test response' } };
-        yield { type: 'message_delta', usage: {} };
-      }
-
-      // Mock createResponse to return the async generator
-      jest.spyOn(client, 'createResponse').mockImplementation(() => {
-        return mockAsyncGenerator();
-      });
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-opus',
-          temperature: 0.7,
-          topK: 10,
-          topP: 0.9,
-        },
-      });
-
-      // Mock getClient to capture the request options
-      let capturedOptions = null;
-      jest.spyOn(client, 'getClient').mockImplementation((options) => {
-        capturedOptions = options;
-        return {};
-      });
-
-      const payload = [{ role: 'user', content: 'Test message' }];
-      await client.sendCompletion(payload, {});
-
-      // Check the options passed to getClient
-      expect(capturedOptions).toHaveProperty('top_k', 10);
-      expect(capturedOptions).toHaveProperty('top_p', 0.9);
-    });
-
-    it('should include top_k and top_p parameters for claude-3-5-sonnet models', async () => {
-      const client = new AnthropicClient('test-api-key');
-
-      // Create a mock async generator function
-      async function* mockAsyncGenerator() {
-        yield { type: 'message_start', message: { usage: {} } };
-        yield { delta: { text: 'Test response' } };
-        yield { type: 'message_delta', usage: {} };
-      }
-
-      // Mock createResponse to return the async generator
-      jest.spyOn(client, 'createResponse').mockImplementation(() => {
-        return mockAsyncGenerator();
-      });
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-5-sonnet',
-          temperature: 0.7,
-          topK: 10,
-          topP: 0.9,
-        },
-      });
-
-      // Mock getClient to capture the request options
-      let capturedOptions = null;
-      jest.spyOn(client, 'getClient').mockImplementation((options) => {
-        capturedOptions = options;
-        return {};
-      });
-
-      const payload = [{ role: 'user', content: 'Test message' }];
-      await client.sendCompletion(payload, {});
-
-      // Check the options passed to getClient
-      expect(capturedOptions).toHaveProperty('top_k', 10);
-      expect(capturedOptions).toHaveProperty('top_p', 0.9);
-    });
-
-    it('should not include top_k and top_p parameters for claude-3-7-sonnet models', async () => {
-      const client = new AnthropicClient('test-api-key');
-
-      // Create a mock async generator function
-      async function* mockAsyncGenerator() {
-        yield { type: 'message_start', message: { usage: {} } };
-        yield { delta: { text: 'Test response' } };
-        yield { type: 'message_delta', usage: {} };
-      }
-
-      // Mock createResponse to return the async generator
-      jest.spyOn(client, 'createResponse').mockImplementation(() => {
-        return mockAsyncGenerator();
-      });
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-7-sonnet',
-          temperature: 0.7,
-          topK: 10,
-          topP: 0.9,
-        },
-      });
-
-      // Mock getClient to capture the request options
-      let capturedOptions = null;
-      jest.spyOn(client, 'getClient').mockImplementation((options) => {
-        capturedOptions = options;
-        return {};
-      });
-
-      const payload = [{ role: 'user', content: 'Test message' }];
-      await client.sendCompletion(payload, {});
-
-      // Check the options passed to getClient
-      expect(capturedOptions).not.toHaveProperty('top_k');
-      expect(capturedOptions).not.toHaveProperty('top_p');
-    });
-
-    it('should not include top_k and top_p parameters for models with decimal notation (claude-3.7)', async () => {
-      const client = new AnthropicClient('test-api-key');
-
-      // Create a mock async generator function
-      async function* mockAsyncGenerator() {
-        yield { type: 'message_start', message: { usage: {} } };
-        yield { delta: { text: 'Test response' } };
-        yield { type: 'message_delta', usage: {} };
-      }
-
-      // Mock createResponse to return the async generator
-      jest.spyOn(client, 'createResponse').mockImplementation(() => {
-        return mockAsyncGenerator();
-      });
-
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.7-sonnet',
-          temperature: 0.7,
-          topK: 10,
-          topP: 0.9,
-        },
-      });
-
-      // Mock getClient to capture the request options
-      let capturedOptions = null;
-      jest.spyOn(client, 'getClient').mockImplementation((options) => {
-        capturedOptions = options;
-        return {};
-      });
-
-      const payload = [{ role: 'user', content: 'Test message' }];
-      await client.sendCompletion(payload, {});
-
-      // Check the options passed to getClient
-      expect(capturedOptions).not.toHaveProperty('top_k');
-      expect(capturedOptions).not.toHaveProperty('top_p');
-    });
-  });
-
-  it('should include top_k and top_p parameters for Claude-3.7 models when thinking is explicitly disabled', async () => {
-    const client = new AnthropicClient('test-api-key', {
-      modelOptions: {
-        model: 'claude-3-7-sonnet',
-        temperature: 0.7,
-        topK: 10,
-        topP: 0.9,
-      },
-      thinking: false,
-    });
-
-    async function* mockAsyncGenerator() {
-      yield { type: 'message_start', message: { usage: {} } };
-      yield { delta: { text: 'Test response' } };
-      yield { type: 'message_delta', usage: {} };
-    }
-
-    jest.spyOn(client, 'createResponse').mockImplementation(() => {
-      return mockAsyncGenerator();
-    });
-
-    let capturedOptions = null;
-    jest.spyOn(client, 'getClient').mockImplementation((options) => {
-      capturedOptions = options;
-      return {};
-    });
-
-    const payload = [{ role: 'user', content: 'Test message' }];
-    await client.sendCompletion(payload, {});
-
-    expect(capturedOptions).toHaveProperty('topK', 10);
-    expect(capturedOptions).toHaveProperty('topP', 0.9);
-
-    client.setOptions({
-      modelOptions: {
-        model: 'claude-3.7-sonnet',
-        temperature: 0.7,
-        topK: 10,
-        topP: 0.9,
-      },
-      thinking: false,
-    });
-
-    await client.sendCompletion(payload, {});
-
-    expect(capturedOptions).toHaveProperty('topK', 10);
-    expect(capturedOptions).toHaveProperty('topP', 0.9);
-  });
-
-  describe('isClaudeLatest', () => {
-    it('should set isClaudeLatest to true for claude-3 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3-sonnet-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(true);
-    });
-
-    it('should set isClaudeLatest to true for claude-3.5 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.5-sonnet-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(true);
-    });
-
-    it('should set isClaudeLatest to true for claude-sonnet-4 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-sonnet-4-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(true);
-    });
-
-    it('should set isClaudeLatest to true for claude-opus-4 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-opus-4-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(true);
-    });
-
-    it('should set isClaudeLatest to true for claude-3.5-haiku models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-3.5-haiku-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(true);
-    });
-
-    it('should set isClaudeLatest to false for claude-2 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-2',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(false);
-    });
-
-    it('should set isClaudeLatest to false for claude-instant models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-instant',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(false);
-    });
-
-    it('should set isClaudeLatest to false for claude-sonnet-3 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-sonnet-3-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(false);
-    });
-
-    it('should set isClaudeLatest to false for claude-opus-3 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-opus-3-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(false);
-    });
-
-    it('should set isClaudeLatest to false for claude-haiku-3 models', () => {
-      const client = new AnthropicClient('test-api-key');
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-haiku-3-20240229',
-        },
-      });
-      expect(client.isClaudeLatest).toBe(false);
-    });
-  });
-
-  describe('configureReasoning', () => {
-    it('should enable thinking for claude-opus-4 and claude-sonnet-4 models', async () => {
-      const client = new AnthropicClient('test-api-key');
-      // Create a mock async generator function
-      async function* mockAsyncGenerator() {
-        yield { type: 'message_start', message: { usage: {} } };
-        yield { delta: { text: 'Test response' } };
-        yield { type: 'message_delta', usage: {} };
-      }
-
-      // Mock createResponse to return the async generator
-      jest.spyOn(client, 'createResponse').mockImplementation(() => {
-        return mockAsyncGenerator();
-      });
-
-      // Test claude-opus-4
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-opus-4-20250514',
-        },
-        thinking: true,
-        thinkingBudget: 2000,
-      });
-
-      let capturedOptions = null;
-      jest.spyOn(client, 'getClient').mockImplementation((options) => {
-        capturedOptions = options;
-        return {};
-      });
-
-      const payload = [{ role: 'user', content: 'Test message' }];
-      await client.sendCompletion(payload, {});
-
-      expect(capturedOptions).toHaveProperty('thinking');
-      expect(capturedOptions.thinking).toEqual({
-        type: 'enabled',
-        budget_tokens: 2000,
-      });
-
-      // Test claude-sonnet-4
-      client.setOptions({
-        modelOptions: {
-          model: 'claude-sonnet-4-20250514',
-        },
-        thinking: true,
-        thinkingBudget: 2000,
-      });
-
-      await client.sendCompletion(payload, {});
-
-      expect(capturedOptions).toHaveProperty('thinking');
-      expect(capturedOptions.thinking).toEqual({
-        type: 'enabled',
-        budget_tokens: 2000,
-      });
-    });
-  });
-});
-
-describe('Claude Model Tests', () => {
-  it('should handle Claude 3 and 4 series models correctly', () => {
-    const client = new AnthropicClient('test-key');
-    // Claude 3 series models
-    const claude3Models = [
-      'claude-3-opus-20240229',
-      'claude-3-sonnet-20240229',
-      'claude-3-haiku-20240307',
-      'claude-3-5-sonnet-20240620',
-      'claude-3-5-haiku-20240620',
-      'claude-3.5-sonnet-20240620',
-      'claude-3.5-haiku-20240620',
-      'claude-3.7-sonnet-20240620',
-      'claude-3.7-haiku-20240620',
-      'anthropic/claude-3-opus-20240229',
-      'claude-3-opus-20240229/anthropic',
-    ];
-
-    // Claude 4 series models
-    const claude4Models = [
-      'claude-sonnet-4-20250514',
-      'claude-opus-4-20250514',
-      'claude-4-sonnet-20250514',
-      'claude-4-opus-20250514',
-      'anthropic/claude-sonnet-4-20250514',
-      'claude-sonnet-4-20250514/anthropic',
-    ];
-
-    // Test Claude 3 series
-    claude3Models.forEach((model) => {
-      client.setOptions({ modelOptions: { model } });
-      expect(
-        /claude-[3-9]/.test(client.modelOptions.model) ||
-          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
-      ).toBe(true);
-    });
-
-    // Test Claude 4 series
-    claude4Models.forEach((model) => {
-      client.setOptions({ modelOptions: { model } });
-      expect(
-        /claude-[3-9]/.test(client.modelOptions.model) ||
-          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
-      ).toBe(true);
-    });
-
-    // Test non-Claude 3/4 models
-    const nonClaudeModels = ['claude-2', 'claude-instant', 'gpt-4', 'gpt-3.5-turbo'];
-
-    nonClaudeModels.forEach((model) => {
-      client.setOptions({ modelOptions: { model } });
-      expect(
-        /claude-[3-9]/.test(client.modelOptions.model) ||
-          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
-      ).toBe(false);
-    });
-  });
 });
--- a/api/app/clients/specs/BaseClient.test.js
+++ b/api/app/clients/specs/BaseClient.test.js
@@ -1,15 +1,7 @@
 const { Constants } = require('librechat-data-provider');
 const { initializeFakeClient } = require('./FakeClient');

-jest.mock('~/db/connect');
-jest.mock('~/server/services/Config', () => ({
-  getAppConfig: jest.fn().mockResolvedValue({
-    // Default app config for tests
-    paths: { uploads: '/tmp' },
-    fileStrategy: 'local',
-    memory: { disabled: false },
-  }),
-}));
+jest.mock('~/lib/db/connectDb');
 jest.mock('~/models', () => ({
  User: jest.fn(),
  Key: jest.fn(),
@@ -38,12 +30,8 @@ jest.mock('~/models', () => ({
  updateFileUsage: jest.fn(),
 }));

-const { getConvo, saveConvo } = require('~/models');
-
-jest.mock('@librechat/agents', () => {
-  const { Providers } = jest.requireActual('@librechat/agents');
+jest.mock('@langchain/openai', () => {
  return {
-    Providers,
    ChatOpenAI: jest.fn().mockImplementation(() => {
      return {};
    }),
@@ -62,7 +50,7 @@ const messageHistory = [
  {
    role: 'user',
    isCreatedByUser: true,
-    text: "What's up",
+    text: 'What\'s up',
    messageId: '3',
    parentMessageId: '2',
  },
@@ -73,7 +61,7 @@ describe('BaseClient', () => {
  const options = {
    // debug: true,
    modelOptions: {
-      model: 'gpt-4o-mini',
+      model: 'gpt-3.5-turbo',
      temperature: 0,
    },
  };
@@ -100,19 +88,6 @@ describe('BaseClient', () => {
    const messages = [{ content: 'Hello' }, { content: 'How are you?' }, { content: 'Goodbye' }];
    const instructions = { content: 'Please respond to the question.' };
    const result = TestClient.addInstructions(messages, instructions);
-    const expected = [
-      { content: 'Please respond to the question.' },
-      { content: 'Hello' },
-      { content: 'How are you?' },
-      { content: 'Goodbye' },
-    ];
-    expect(result).toEqual(expected);
-  });
-
-  test('returns the input messages with instructions properly added when addInstructions() with legacy flag', () => {
-    const messages = [{ content: 'Hello' }, { content: 'How are you?' }, { content: 'Goodbye' }];
-    const instructions = { content: 'Please respond to the question.' };
-    const result = TestClient.addInstructions(messages, instructions, true);
    const expected = [
      { content: 'Hello' },
      { content: 'How are you?' },
@@ -171,10 +146,10 @@ describe('BaseClient', () => {
      expectedMessagesToRefine?.[expectedMessagesToRefine.length - 1] ?? {};
    const expectedIndex = messages.findIndex((msg) => msg.content === lastExpectedMessage?.content);

-    const result = await TestClient.getMessagesWithinTokenLimit({ messages });
+    const result = await TestClient.getMessagesWithinTokenLimit(messages);

    expect(result.context).toEqual(expectedContext);
-    expect(result.messagesToRefine.length - 1).toEqual(expectedIndex);
+    expect(result.summaryIndex).toEqual(expectedIndex);
    expect(result.remainingContextTokens).toBe(expectedRemainingContextTokens);
    expect(result.messagesToRefine).toEqual(expectedMessagesToRefine);
  });
@@ -207,14 +182,74 @@ describe('BaseClient', () => {
      expectedMessagesToRefine?.[expectedMessagesToRefine.length - 1] ?? {};
    const expectedIndex = messages.findIndex((msg) => msg.content === lastExpectedMessage?.content);

-    const result = await TestClient.getMessagesWithinTokenLimit({ messages });
+    const result = await TestClient.getMessagesWithinTokenLimit(messages);

    expect(result.context).toEqual(expectedContext);
-    expect(result.messagesToRefine.length - 1).toEqual(expectedIndex);
+    expect(result.summaryIndex).toEqual(expectedIndex);
    expect(result.remainingContextTokens).toBe(expectedRemainingContextTokens);
    expect(result.messagesToRefine).toEqual(expectedMessagesToRefine);
  });

+  test('handles context strategy correctly in handleContextStrategy()', async () => {
+    TestClient.addInstructions = jest
+      .fn()
+      .mockReturnValue([
+        { content: 'Hello' },
+        { content: 'How can I help you?' },
+        { content: 'Please provide more details.' },
+        { content: 'I can assist you with that.' },
+      ]);
+    TestClient.getMessagesWithinTokenLimit = jest.fn().mockReturnValue({
+      context: [
+        { content: 'How can I help you?' },
+        { content: 'Please provide more details.' },
+        { content: 'I can assist you with that.' },
+      ],
+      remainingContextTokens: 80,
+      messagesToRefine: [{ content: 'Hello' }],
+      summaryIndex: 3,
+    });
+
+    TestClient.getTokenCount = jest.fn().mockReturnValue(40);
+
+    const instructions = { content: 'Please provide more details.' };
+    const orderedMessages = [
+      { content: 'Hello' },
+      { content: 'How can I help you?' },
+      { content: 'Please provide more details.' },
+      { content: 'I can assist you with that.' },
+    ];
+    const formattedMessages = [
+      { content: 'Hello' },
+      { content: 'How can I help you?' },
+      { content: 'Please provide more details.' },
+      { content: 'I can assist you with that.' },
+    ];
+    const expectedResult = {
+      payload: [
+        {
+          role: 'system',
+          content: 'Refined answer',
+        },
+        { content: 'How can I help you?' },
+        { content: 'Please provide more details.' },
+        { content: 'I can assist you with that.' },
+      ],
+      promptTokens: expect.any(Number),
+      tokenCountMap: {},
+      messages: expect.any(Array),
+    };
+
+    TestClient.shouldSummarize = true;
+    const result = await TestClient.handleContextStrategy({
+      instructions,
+      orderedMessages,
+      formattedMessages,
+    });
+
+    expect(result).toEqual(expectedResult);
+  });
+
  describe('getMessagesForConversation', () => {
    it('should return an empty array if the parentMessageId does not exist', () => {
      const result = TestClient.constructor.getMessagesForConversation({
@@ -430,46 +465,6 @@ describe('BaseClient', () => {
      expect(response).toEqual(expectedResult);
    });

-    test('should replace responseMessageId with new UUID when isRegenerate is true and messageId ends with underscore', async () => {
-      const mockCrypto = require('crypto');
-      const newUUID = 'new-uuid-1234';
-      jest.spyOn(mockCrypto, 'randomUUID').mockReturnValue(newUUID);
-
-      const opts = {
-        isRegenerate: true,
-        responseMessageId: 'existing-message-id_',
-      };
-
-      await TestClient.setMessageOptions(opts);
-
-      expect(TestClient.responseMessageId).toBe(newUUID);
-      expect(TestClient.responseMessageId).not.toBe('existing-message-id_');
-
-      mockCrypto.randomUUID.mockRestore();
-    });
-
-    test('should not replace responseMessageId when isRegenerate is false', async () => {
-      const opts = {
-        isRegenerate: false,
-        responseMessageId: 'existing-message-id_',
-      };
-
-      await TestClient.setMessageOptions(opts);
-
-      expect(TestClient.responseMessageId).toBe('existing-message-id_');
-    });
-
-    test('should not replace responseMessageId when it does not end with underscore', async () => {
-      const opts = {
-        isRegenerate: true,
-        responseMessageId: 'existing-message-id',
-      };
-
-      await TestClient.setMessageOptions(opts);
-
-      expect(TestClient.responseMessageId).toBe('existing-message-id');
-    });
-
    test('sendMessage should work with provided conversationId and parentMessageId', async () => {
      const userMessage = 'Second message in the conversation';
      const opts = {
@@ -506,7 +501,7 @@ describe('BaseClient', () => {

      const chatMessages2 = await TestClient.loadHistory(conversationId, '3');
      expect(TestClient.currentMessages).toHaveLength(3);
-      expect(chatMessages2[chatMessages2.length - 1].text).toEqual("What's up");
+      expect(chatMessages2[chatMessages2.length - 1].text).toEqual('What\'s up');
    });

    /* Most of the new sendMessage logic revolving around edited/continued AI messages
@@ -587,18 +582,15 @@ describe('BaseClient', () => {
      expect(onStart).toHaveBeenCalledWith(
        expect.objectContaining({ text: 'Hello, world!' }),
        expect.any(String),
-        /** `isNewConvo` */
-        true,
      );
    });

    test('saveMessageToDatabase is called with the correct arguments', async () => {
      const saveOptions = TestClient.getSaveOptions();
-      const user = {};
+      const user = {}; // Mock user
      const opts = { user };
-      const saveSpy = jest.spyOn(TestClient, 'saveMessageToDatabase');
      await TestClient.sendMessage('Hello, world!', opts);
-      expect(saveSpy).toHaveBeenCalledWith(
+      expect(TestClient.saveMessageToDatabase).toHaveBeenCalledWith(
        expect.objectContaining({
          sender: expect.any(String),
          text: expect.any(String),
@@ -612,157 +604,6 @@ describe('BaseClient', () => {
      );
    });

-    test('should handle existing conversation when getConvo retrieves one', async () => {
-      const existingConvo = {
-        conversationId: 'existing-convo-id',
-        endpoint: 'openai',
-        endpointType: 'openai',
-        model: 'gpt-3.5-turbo',
-        messages: [
-          { role: 'user', content: 'Existing message 1' },
-          { role: 'assistant', content: 'Existing response 1' },
-        ],
-        temperature: 1,
-      };
-
-      const { temperature: _temp, ...newConvo } = existingConvo;
-
-      const user = {
-        id: 'user-id',
-      };
-
-      getConvo.mockResolvedValue(existingConvo);
-      saveConvo.mockResolvedValue(newConvo);
-
-      TestClient = initializeFakeClient(
-        apiKey,
-        {
-          ...options,
-          req: {
-            user,
-          },
-        },
-        [],
-      );
-
-      const saveSpy = jest.spyOn(TestClient, 'saveMessageToDatabase');
-
-      const newMessage = 'New message in existing conversation';
-      const response = await TestClient.sendMessage(newMessage, {
-        user,
-        conversationId: existingConvo.conversationId,
-      });
-
-      expect(getConvo).toHaveBeenCalledWith(user.id, existingConvo.conversationId);
-      expect(TestClient.conversationId).toBe(existingConvo.conversationId);
-      expect(response.conversationId).toBe(existingConvo.conversationId);
-      expect(TestClient.fetchedConvo).toBe(true);
-
-      expect(saveSpy).toHaveBeenCalledWith(
-        expect.objectContaining({
-          conversationId: existingConvo.conversationId,
-          text: newMessage,
-        }),
-        expect.any(Object),
-        expect.any(Object),
-      );
-
-      expect(saveConvo).toHaveBeenCalledTimes(2);
-      expect(saveConvo).toHaveBeenCalledWith(
-        expect.any(Object),
-        expect.objectContaining({
-          conversationId: existingConvo.conversationId,
-        }),
-        expect.objectContaining({
-          context: 'api/app/clients/BaseClient.js - saveMessageToDatabase #saveConvo',
-          unsetFields: {
-            temperature: 1,
-          },
-        }),
-      );
-
-      await TestClient.sendMessage('Another message', {
-        conversationId: existingConvo.conversationId,
-      });
-      expect(getConvo).toHaveBeenCalledTimes(1);
-    });
-
-    test('should correctly handle existing conversation and unset fields appropriately', async () => {
-      const existingConvo = {
-        conversationId: 'existing-convo-id',
-        endpoint: 'openai',
-        endpointType: 'openai',
-        model: 'gpt-3.5-turbo',
-        messages: [
-          { role: 'user', content: 'Existing message 1' },
-          { role: 'assistant', content: 'Existing response 1' },
-        ],
-        title: 'Existing Conversation',
-        someExistingField: 'existingValue',
-        anotherExistingField: 'anotherValue',
-        temperature: 0.7,
-        modelLabel: 'GPT-3.5',
-      };
-
-      getConvo.mockResolvedValue(existingConvo);
-      saveConvo.mockResolvedValue(existingConvo);
-
-      TestClient = initializeFakeClient(
-        apiKey,
-        {
-          ...options,
-          modelOptions: {
-            model: 'gpt-4',
-            temperature: 0.5,
-          },
-        },
-        [],
-      );
-
-      const newMessage = 'New message in existing conversation';
-      await TestClient.sendMessage(newMessage, {
-        conversationId: existingConvo.conversationId,
-      });
-
-      expect(saveConvo).toHaveBeenCalledTimes(2);
-
-      const saveConvoCall = saveConvo.mock.calls[0];
-      const [, savedFields, saveOptions] = saveConvoCall;
-
-      // Instead of checking all excludedKeys, we'll just check specific fields
-      // that we know should be excluded
-      expect(savedFields).not.toHaveProperty('messages');
-      expect(savedFields).not.toHaveProperty('title');
-
-      // Only check that someExistingField is in unsetFields
-      expect(saveOptions.unsetFields).toHaveProperty('someExistingField', 1);
-
-      // Mock saveConvo to return the expected fields
-      saveConvo.mockImplementation((req, fields) => {
-        return Promise.resolve({
-          ...fields,
-          endpoint: 'openai',
-          endpointType: 'openai',
-          model: 'gpt-4',
-          temperature: 0.5,
-        });
-      });
-
-      // Only check the conversationId since that's the only field we can be sure about
-      expect(savedFields).toHaveProperty('conversationId', 'existing-convo-id');
-
-      expect(TestClient.fetchedConvo).toBe(true);
-
-      await TestClient.sendMessage('Another message', {
-        conversationId: existingConvo.conversationId,
-      });
-
-      expect(getConvo).toHaveBeenCalledTimes(1);
-
-      const secondSaveConvoCall = saveConvo.mock.calls[1];
-      expect(secondSaveConvoCall[2]).toHaveProperty('unsetFields', {});
-    });
-
    test('sendCompletion is called with the correct arguments', async () => {
      const payload = {}; // Mock payload
      TestClient.buildMessages.mockReturnValue({ prompt: payload, tokenCountMap: null });
@@ -774,9 +615,9 @@ describe('BaseClient', () => {
    test('getTokenCount for response is called with the correct arguments', async () => {
      const tokenCountMap = {}; // Mock tokenCountMap
      TestClient.buildMessages.mockReturnValue({ prompt: [], tokenCountMap });
-      TestClient.getTokenCountForResponse = jest.fn();
+      TestClient.getTokenCount = jest.fn();
      const response = await TestClient.sendMessage('Hello, world!', {});
-      expect(TestClient.getTokenCountForResponse).toHaveBeenCalledWith(response);
+      expect(TestClient.getTokenCount).toHaveBeenCalledWith(response.text);
    });

    test('returns an object with the correct shape', async () => {
@@ -820,112 +661,4 @@ describe('BaseClient', () => {
      expect(calls[1][0].isCreatedByUser).toBe(false); // Second call should be for response message
    });
  });
-
-  describe('getMessagesWithinTokenLimit with instructions', () => {
-    test('should always include instructions when present', async () => {
-      TestClient.maxContextTokens = 50;
-      const instructions = {
-        role: 'system',
-        content: 'System instructions',
-        tokenCount: 20,
-      };
-
-      const messages = [
-        instructions,
-        { role: 'user', content: 'Hello', tokenCount: 10 },
-        { role: 'assistant', content: 'Hi there', tokenCount: 15 },
-      ];
-
-      const result = await TestClient.getMessagesWithinTokenLimit({
-        messages,
-        instructions,
-      });
-
-      expect(result.context[0]).toBe(instructions);
-      expect(result.remainingContextTokens).toBe(2);
-    });
-
-    test('should handle case when messages exceed limit but instructions must be preserved', async () => {
-      TestClient.maxContextTokens = 30;
-      const instructions = {
-        role: 'system',
-        content: 'System instructions',
-        tokenCount: 20,
-      };
-
-      const messages = [
-        instructions,
-        { role: 'user', content: 'Hello', tokenCount: 10 },
-        { role: 'assistant', content: 'Hi there', tokenCount: 15 },
-      ];
-
-      const result = await TestClient.getMessagesWithinTokenLimit({
-        messages,
-        instructions,
-      });
-
-      // Should only include instructions and the last message that fits
-      expect(result.context).toHaveLength(1);
-      expect(result.context[0].content).toBe(instructions.content);
-      expect(result.messagesToRefine).toHaveLength(2);
-      expect(result.remainingContextTokens).toBe(7); // 30 - 20 - 3 (assistant label)
-    });
-
-    test('should work correctly without instructions (1/2)', async () => {
-      TestClient.maxContextTokens = 50;
-      const messages = [
-        { role: 'user', content: 'Hello', tokenCount: 10 },
-        { role: 'assistant', content: 'Hi there', tokenCount: 15 },
-      ];
-
-      const result = await TestClient.getMessagesWithinTokenLimit({
-        messages,
-      });
-
-      expect(result.context).toHaveLength(2);
-      expect(result.remainingContextTokens).toBe(22); // 50 - 10 - 15 - 3(assistant label)
-      expect(result.messagesToRefine).toHaveLength(0);
-    });
-
-    test('should work correctly without instructions (2/2)', async () => {
-      TestClient.maxContextTokens = 30;
-      const messages = [
-        { role: 'user', content: 'Hello', tokenCount: 10 },
-        { role: 'assistant', content: 'Hi there', tokenCount: 20 },
-      ];
-
-      const result = await TestClient.getMessagesWithinTokenLimit({
-        messages,
-      });
-
-      expect(result.context).toHaveLength(1);
-      expect(result.remainingContextTokens).toBe(7);
-      expect(result.messagesToRefine).toHaveLength(1);
-    });
-
-    test('should handle case when only instructions fit within limit', async () => {
-      TestClient.maxContextTokens = 25;
-      const instructions = {
-        role: 'system',
-        content: 'System instructions',
-        tokenCount: 20,
-      };
-
-      const messages = [
-        instructions,
-        { role: 'user', content: 'Hello', tokenCount: 10 },
-        { role: 'assistant', content: 'Hi there', tokenCount: 15 },
-      ];
-
-      const result = await TestClient.getMessagesWithinTokenLimit({
-        messages,
-        instructions,
-      });
-
-      expect(result.context).toHaveLength(1);
-      expect(result.context[0]).toBe(instructions);
-      expect(result.messagesToRefine).toHaveLength(2);
-      expect(result.remainingContextTokens).toBe(2); // 25 - 20 - 3(assistant label)
-    });
-  });
 });
--- a/api/app/clients/specs/FakeClient.js
+++ b/api/app/clients/specs/FakeClient.js
@@ -1,5 +1,5 @@
-const { getModelMaxTokens } = require('@librechat/api');
 const BaseClient = require('../BaseClient');
+const { getModelMaxTokens } = require('../../../utils');

 class FakeClient extends BaseClient {
  constructor(apiKey, options = {}) {
@@ -56,6 +56,7 @@ const initializeFakeClient = (apiKey, options, fakeMessages) => {
  let TestClient = new FakeClient(apiKey);
  TestClient.options = options;
  TestClient.abortController = { abort: jest.fn() };
+  TestClient.saveMessageToDatabase = jest.fn();
  TestClient.loadHistory = jest
    .fn()
    .mockImplementation((conversationId, parentMessageId = null) => {
@@ -85,6 +86,7 @@ const initializeFakeClient = (apiKey, options, fakeMessages) => {
    return 'Mock response text';
  });

+  // eslint-disable-next-line no-unused-vars
  TestClient.getCompletion = jest.fn().mockImplementation(async (..._args) => {
    return {
      choices: [
--- a/api/app/clients/specs/OpenAIClient.test.js
+++ b/api/app/clients/specs/OpenAIClient.test.js
@@ -1,11 +1,11 @@
-jest.mock('~/cache/getLogStores');
 require('dotenv').config();
+const OpenAI = require('openai');
 const { fetchEventSource } = require('@waylaidwanderer/fetch-event-source');
-const getLogStores = require('~/cache/getLogStores');
+const { genAzureChatCompletion } = require('~/utils/azureUtils');
 const OpenAIClient = require('../OpenAIClient');
 jest.mock('meilisearch');

-jest.mock('~/db/connect');
+jest.mock('~/lib/db/connectDb');
 jest.mock('~/models', () => ({
  User: jest.fn(),
  Key: jest.fn(),
@@ -34,21 +34,19 @@ jest.mock('~/models', () => ({
  updateFileUsage: jest.fn(),
 }));

-// Import the actual module but mock specific parts
-const agents = jest.requireActual('@librechat/agents');
-const { CustomOpenAIClient } = agents;
-
-// Also mock ChatOpenAI to prevent real API calls
-agents.ChatOpenAI = jest.fn().mockImplementation(() => {
-  return {};
-});
-agents.AzureChatOpenAI = jest.fn().mockImplementation(() => {
-  return {};
+jest.mock('@langchain/openai', () => {
+  return {
+    ChatOpenAI: jest.fn().mockImplementation(() => {
+      return {};
+    }),
+  };
 });

-// Mock only the CustomOpenAIClient constructor
-jest.spyOn(CustomOpenAIClient, 'constructor').mockImplementation(function (...options) {
-  return new CustomOpenAIClient(...options);
+jest.mock('openai');
+
+jest.spyOn(OpenAI, 'constructor').mockImplementation(function (...options) {
+  // We can add additional logic here if needed
+  return new OpenAI(...options);
 });

 const finalChatCompletion = jest.fn().mockResolvedValue({
@@ -120,13 +118,7 @@ const create = jest.fn().mockResolvedValue({
  ],
 });

-// Mock the implementation of CustomOpenAIClient instances
-jest.spyOn(CustomOpenAIClient.prototype, 'constructor').mockImplementation(function () {
-  return this;
-});
-
-// Create a mock for the CustomOpenAIClient class
-const mockCustomOpenAIClient = jest.fn().mockImplementation(() => ({
+OpenAI.mockImplementation(() => ({
  beta: {
    chat: {
      completions: {
@@ -141,17 +133,8 @@ const mockCustomOpenAIClient = jest.fn().mockImplementation(() => ({
  },
 }));

-CustomOpenAIClient.mockImplementation = mockCustomOpenAIClient;
-
 describe('OpenAIClient', () => {
-  beforeEach(() => {
-    const mockCache = {
-      get: jest.fn().mockResolvedValue({}),
-      set: jest.fn(),
-    };
-    getLogStores.mockReturnValue(mockCache);
-  });
-  let client;
+  let client, client2;
  const model = 'gpt-4';
  const parentMessageId = '1';
  const messages = [
@@ -193,6 +176,7 @@ describe('OpenAIClient', () => {
  beforeEach(() => {
    const options = { ...defaultOptions };
    client = new OpenAIClient('test-api-key', options);
+    client2 = new OpenAIClient('test-api-key', options);
    client.summarizeMessages = jest.fn().mockResolvedValue({
      role: 'assistant',
      content: 'Refined answer',
@@ -201,6 +185,7 @@ describe('OpenAIClient', () => {
    client.buildPrompt = jest
      .fn()
      .mockResolvedValue({ prompt: messages.map((m) => m.text).join('\n') });
+    client.constructor.freeAndResetAllEncoders();
    client.getMessages = jest.fn().mockResolvedValue([]);
  });

@@ -211,6 +196,14 @@ describe('OpenAIClient', () => {
      expect(client.modelOptions.temperature).toBe(0.7);
    });

+    it('should set apiKey and useOpenRouter if OPENROUTER_API_KEY is present', () => {
+      process.env.OPENROUTER_API_KEY = 'openrouter-key';
+      client.setOptions({});
+      expect(client.apiKey).toBe('openrouter-key');
+      expect(client.useOpenRouter).toBe(true);
+      delete process.env.OPENROUTER_API_KEY; // Cleanup
+    });
+
    it('should set FORCE_PROMPT based on OPENAI_FORCE_PROMPT or reverseProxyUrl', () => {
      process.env.OPENAI_FORCE_PROMPT = 'true';
      client.setOptions({});
@@ -228,7 +221,7 @@ describe('OpenAIClient', () => {

    it('should set isChatCompletion based on useOpenRouter, reverseProxyUrl, or model', () => {
      client.setOptions({ reverseProxyUrl: null });
-      // true by default since default model will be gpt-4o-mini
+      // true by default since default model will be gpt-3.5-turbo
      expect(client.isChatCompletion).toBe(true);
      client.isChatCompletion = undefined;

@@ -237,7 +230,7 @@ describe('OpenAIClient', () => {
      expect(client.isChatCompletion).toBe(false);
      client.isChatCompletion = undefined;

-      client.setOptions({ modelOptions: { model: 'gpt-4o-mini' }, reverseProxyUrl: null });
+      client.setOptions({ modelOptions: { model: 'gpt-3.5-turbo' }, reverseProxyUrl: null });
      expect(client.isChatCompletion).toBe(true);
    });

@@ -342,18 +335,83 @@ describe('OpenAIClient', () => {
    });
  });

+  describe('selectTokenizer', () => {
+    it('should get the correct tokenizer based on the instance state', () => {
+      const tokenizer = client.selectTokenizer();
+      expect(tokenizer).toBeDefined();
+    });
+  });
+
+  describe('freeAllTokenizers', () => {
+    it('should free all tokenizers', () => {
+      // Create a tokenizer
+      const tokenizer = client.selectTokenizer();
+
+      // Mock 'free' method on the tokenizer
+      tokenizer.free = jest.fn();
+
+      client.constructor.freeAndResetAllEncoders();
+
+      // Check if 'free' method has been called on the tokenizer
+      expect(tokenizer.free).toHaveBeenCalled();
+    });
+  });
+
  describe('getTokenCount', () => {
    it('should return the correct token count', () => {
      const count = client.getTokenCount('Hello, world!');
      expect(count).toBeGreaterThan(0);
    });
+
+    it('should reset the encoder and count when count reaches 25', () => {
+      const freeAndResetEncoderSpy = jest.spyOn(client.constructor, 'freeAndResetAllEncoders');
+
+      // Call getTokenCount 25 times
+      for (let i = 0; i < 25; i++) {
+        client.getTokenCount('test text');
+      }
+
+      expect(freeAndResetEncoderSpy).toHaveBeenCalled();
+    });
+
+    it('should not reset the encoder and count when count is less than 25', () => {
+      const freeAndResetEncoderSpy = jest.spyOn(client.constructor, 'freeAndResetAllEncoders');
+      freeAndResetEncoderSpy.mockClear();
+
+      // Call getTokenCount 24 times
+      for (let i = 0; i < 24; i++) {
+        client.getTokenCount('test text');
+      }
+
+      expect(freeAndResetEncoderSpy).not.toHaveBeenCalled();
+    });
+
+    it('should handle errors and reset the encoder', () => {
+      const freeAndResetEncoderSpy = jest.spyOn(client.constructor, 'freeAndResetAllEncoders');
+
+      // Mock encode function to throw an error
+      client.selectTokenizer().encode = jest.fn().mockImplementation(() => {
+        throw new Error('Test error');
+      });
+
+      client.getTokenCount('test text');
+
+      expect(freeAndResetEncoderSpy).toHaveBeenCalled();
+    });
+
+    it('should not throw null pointer error when freeing the same encoder twice', () => {
+      client.constructor.freeAndResetAllEncoders();
+      client2.constructor.freeAndResetAllEncoders();
+
+      const count = client2.getTokenCount('test text');
+      expect(count).toBeGreaterThan(0);
+    });
  });

  describe('getSaveOptions', () => {
    it('should return the correct save options', () => {
      const options = client.getSaveOptions();
      expect(options).toHaveProperty('chatGptLabel');
-      expect(options).toHaveProperty('modelLabel');
      expect(options).toHaveProperty('promptPrefix');
    });
  });
@@ -462,17 +520,17 @@ describe('OpenAIClient', () => {
        role: 'system',
        name: 'example_user',
        content:
-          "Let's circle back when we have more bandwidth to touch base on opportunities for increased leverage.",
+          'Let\'s circle back when we have more bandwidth to touch base on opportunities for increased leverage.',
      },
      {
        role: 'system',
        name: 'example_assistant',
-        content: "Let's talk later when we're less busy about how to do better.",
+        content: 'Let\'s talk later when we\'re less busy about how to do better.',
      },
      {
        role: 'user',
        content:
-          "This late pivot means we don't have time to boil the ocean for the client deliverable.",
+          'This late pivot means we don\'t have time to boil the ocean for the client deliverable.',
      },
    ];

@@ -489,6 +547,7 @@ describe('OpenAIClient', () => {
    testCases.forEach((testCase) => {
      it(`should return ${testCase.expected} tokens for model ${testCase.model}`, () => {
        client.modelOptions.model = testCase.model;
+        client.selectTokenizer();
        // 3 tokens for assistant label
        let totalTokens = 3;
        for (let message of example_messages) {
@@ -522,6 +581,7 @@ describe('OpenAIClient', () => {

    it(`should return ${expectedTokens} tokens for model ${visionModel} (Vision Request)`, () => {
      client.modelOptions.model = visionModel;
+      client.selectTokenizer();
      // 3 tokens for assistant label
      let totalTokens = 3;
      for (let message of vision_request) {
@@ -531,6 +591,80 @@ describe('OpenAIClient', () => {
    });
  });

+  describe('sendMessage/getCompletion/chatCompletion', () => {
+    afterEach(() => {
+      delete process.env.AZURE_OPENAI_DEFAULT_MODEL;
+      delete process.env.AZURE_USE_MODEL_AS_DEPLOYMENT_NAME;
+      delete process.env.OPENROUTER_API_KEY;
+    });
+
+    it('should call getCompletion and fetchEventSource when using a text/instruct model', async () => {
+      const model = 'text-davinci-003';
+      const onProgress = jest.fn().mockImplementation(() => ({}));
+
+      const testClient = new OpenAIClient('test-api-key', {
+        ...defaultOptions,
+        modelOptions: { model },
+      });
+
+      const getCompletion = jest.spyOn(testClient, 'getCompletion');
+      await testClient.sendMessage('Hi mom!', { onProgress });
+
+      expect(getCompletion).toHaveBeenCalled();
+      expect(getCompletion.mock.calls.length).toBe(1);
+
+      expect(getCompletion.mock.calls[0][0]).toBe('||>User:\nHi mom!\n||>Assistant:\n');
+
+      expect(fetchEventSource).toHaveBeenCalled();
+      expect(fetchEventSource.mock.calls.length).toBe(1);
+
+      // Check if the first argument (url) is correct
+      const firstCallArgs = fetchEventSource.mock.calls[0];
+
+      const expectedURL = 'https://api.openai.com/v1/completions';
+      expect(firstCallArgs[0]).toBe(expectedURL);
+
+      const requestBody = JSON.parse(firstCallArgs[1].body);
+      expect(requestBody).toHaveProperty('model');
+      expect(requestBody.model).toBe(model);
+    });
+
+    it('[Azure OpenAI] should call chatCompletion and OpenAI.stream with correct args', async () => {
+      // Set a default model
+      process.env.AZURE_OPENAI_DEFAULT_MODEL = 'gpt4-turbo';
+
+      const onProgress = jest.fn().mockImplementation(() => ({}));
+      client.azure = defaultAzureOptions;
+      const chatCompletion = jest.spyOn(client, 'chatCompletion');
+      await client.sendMessage('Hi mom!', {
+        replaceOptions: true,
+        ...defaultOptions,
+        modelOptions: { model: 'gpt4-turbo', stream: true },
+        onProgress,
+        azure: defaultAzureOptions,
+      });
+
+      expect(chatCompletion).toHaveBeenCalled();
+      expect(chatCompletion.mock.calls.length).toBe(1);
+
+      const chatCompletionArgs = chatCompletion.mock.calls[0][0];
+      const { payload } = chatCompletionArgs;
+
+      expect(payload[0].role).toBe('user');
+      expect(payload[0].content).toBe('Hi mom!');
+
+      // Azure OpenAI does not use the model property, and will error if it's passed
+      // This check ensures the model property is not present
+      const streamArgs = stream.mock.calls[0][0];
+      expect(streamArgs).not.toHaveProperty('model');
+
+      // Check if the baseURL is correct
+      const constructorArgs = OpenAI.mock.calls[0][0];
+      const expectedURL = genAzureChatCompletion(defaultAzureOptions).split('/chat')[0];
+      expect(constructorArgs.baseURL).toBe(expectedURL);
+    });
+  });
+
  describe('checkVisionRequest functionality', () => {
    let client;
    const attachments = [{ type: 'image/png' }];
--- a/api/app/clients/specs/PluginsClient.test.js
+++ b/api/app/clients/specs/PluginsClient.test.js
@@ -0,0 +1,314 @@
+const crypto = require('crypto');
+const { Constants } = require('librechat-data-provider');
+const { HumanMessage, AIMessage } = require('@langchain/core/messages');
+const PluginsClient = require('../PluginsClient');
+
+jest.mock('~/lib/db/connectDb');
+jest.mock('~/models/Conversation', () => {
+  return function () {
+    return {
+      save: jest.fn(),
+      deleteConvos: jest.fn(),
+    };
+  };
+});
+
+const defaultAzureOptions = {
+  azureOpenAIApiInstanceName: 'your-instance-name',
+  azureOpenAIApiDeploymentName: 'your-deployment-name',
+  azureOpenAIApiVersion: '2020-07-01-preview',
+};
+
+describe('PluginsClient', () => {
+  let TestAgent;
+  let options = {
+    tools: [],
+    modelOptions: {
+      model: 'gpt-3.5-turbo',
+      temperature: 0,
+      max_tokens: 2,
+    },
+    agentOptions: {
+      model: 'gpt-3.5-turbo',
+    },
+  };
+  let parentMessageId;
+  let conversationId;
+  const fakeMessages = [];
+  const userMessage = 'Hello, ChatGPT!';
+  const apiKey = 'fake-api-key';
+
+  beforeEach(() => {
+    TestAgent = new PluginsClient(apiKey, options);
+    TestAgent.loadHistory = jest
+      .fn()
+      .mockImplementation((conversationId, parentMessageId = null) => {
+        if (!conversationId) {
+          TestAgent.currentMessages = [];
+          return Promise.resolve([]);
+        }
+
+        const orderedMessages = TestAgent.constructor.getMessagesForConversation({
+          messages: fakeMessages,
+          parentMessageId,
+        });
+
+        const chatMessages = orderedMessages.map((msg) =>
+          msg?.isCreatedByUser || msg?.role?.toLowerCase() === 'user'
+            ? new HumanMessage(msg.text)
+            : new AIMessage(msg.text),
+        );
+
+        TestAgent.currentMessages = orderedMessages;
+        return Promise.resolve(chatMessages);
+      });
+    TestAgent.sendMessage = jest.fn().mockImplementation(async (message, opts = {}) => {
+      if (opts && typeof opts === 'object') {
+        TestAgent.setOptions(opts);
+      }
+      const conversationId = opts.conversationId || crypto.randomUUID();
+      const parentMessageId = opts.parentMessageId || Constants.NO_PARENT;
+      const userMessageId = opts.overrideParentMessageId || crypto.randomUUID();
+      this.pastMessages = await TestAgent.loadHistory(
+        conversationId,
+        TestAgent.options?.parentMessageId,
+      );
+
+      const userMessage = {
+        text: message,
+        sender: 'ChatGPT',
+        isCreatedByUser: true,
+        messageId: userMessageId,
+        parentMessageId,
+        conversationId,
+      };
+
+      const response = {
+        sender: 'ChatGPT',
+        text: 'Hello, User!',
+        isCreatedByUser: false,
+        messageId: crypto.randomUUID(),
+        parentMessageId: userMessage.messageId,
+        conversationId,
+      };
+
+      fakeMessages.push(userMessage);
+      fakeMessages.push(response);
+      return response;
+    });
+  });
+
+  test('initializes PluginsClient without crashing', () => {
+    expect(TestAgent).toBeInstanceOf(PluginsClient);
+  });
+
+  test('check setOptions function', () => {
+    expect(TestAgent.agentIsGpt3).toBe(true);
+  });
+
+  describe('sendMessage', () => {
+    test('sendMessage should return a response message', async () => {
+      const expectedResult = expect.objectContaining({
+        sender: 'ChatGPT',
+        text: expect.any(String),
+        isCreatedByUser: false,
+        messageId: expect.any(String),
+        parentMessageId: expect.any(String),
+        conversationId: expect.any(String),
+      });
+
+      const response = await TestAgent.sendMessage(userMessage);
+      parentMessageId = response.messageId;
+      conversationId = response.conversationId;
+      expect(response).toEqual(expectedResult);
+    });
+
+    test('sendMessage should work with provided conversationId and parentMessageId', async () => {
+      const userMessage = 'Second message in the conversation';
+      const opts = {
+        conversationId,
+        parentMessageId,
+      };
+
+      const expectedResult = expect.objectContaining({
+        sender: 'ChatGPT',
+        text: expect.any(String),
+        isCreatedByUser: false,
+        messageId: expect.any(String),
+        parentMessageId: expect.any(String),
+        conversationId: opts.conversationId,
+      });
+
+      const response = await TestAgent.sendMessage(userMessage, opts);
+      parentMessageId = response.messageId;
+      expect(response.conversationId).toEqual(conversationId);
+      expect(response).toEqual(expectedResult);
+    });
+
+    test('should return chat history', async () => {
+      const chatMessages = await TestAgent.loadHistory(conversationId, parentMessageId);
+      expect(TestAgent.currentMessages).toHaveLength(4);
+      expect(chatMessages[0].text).toEqual(userMessage);
+    });
+  });
+
+  describe('getFunctionModelName', () => {
+    let client;
+
+    beforeEach(() => {
+      client = new PluginsClient('dummy_api_key');
+    });
+
+    test('should return the input when it includes a dash followed by four digits', () => {
+      expect(client.getFunctionModelName('-1234')).toBe('-1234');
+      expect(client.getFunctionModelName('gpt-4-5678-preview')).toBe('gpt-4-5678-preview');
+    });
+
+    test('should return the input for all function-capable models (`0613` models and above)', () => {
+      expect(client.getFunctionModelName('gpt-4-0613')).toBe('gpt-4-0613');
+      expect(client.getFunctionModelName('gpt-4-32k-0613')).toBe('gpt-4-32k-0613');
+      expect(client.getFunctionModelName('gpt-3.5-turbo-0613')).toBe('gpt-3.5-turbo-0613');
+      expect(client.getFunctionModelName('gpt-3.5-turbo-16k-0613')).toBe('gpt-3.5-turbo-16k-0613');
+      expect(client.getFunctionModelName('gpt-3.5-turbo-1106')).toBe('gpt-3.5-turbo-1106');
+      expect(client.getFunctionModelName('gpt-4-1106-preview')).toBe('gpt-4-1106-preview');
+      expect(client.getFunctionModelName('gpt-4-1106')).toBe('gpt-4-1106');
+    });
+
+    test('should return the corresponding model if input is non-function capable (`0314` models)', () => {
+      expect(client.getFunctionModelName('gpt-4-0314')).toBe('gpt-4');
+      expect(client.getFunctionModelName('gpt-4-32k-0314')).toBe('gpt-4');
+      expect(client.getFunctionModelName('gpt-3.5-turbo-0314')).toBe('gpt-3.5-turbo');
+      expect(client.getFunctionModelName('gpt-3.5-turbo-16k-0314')).toBe('gpt-3.5-turbo');
+    });
+
+    test('should return "gpt-3.5-turbo" when the input includes "gpt-3.5-turbo"', () => {
+      expect(client.getFunctionModelName('test gpt-3.5-turbo model')).toBe('gpt-3.5-turbo');
+    });
+
+    test('should return "gpt-4" when the input includes "gpt-4"', () => {
+      expect(client.getFunctionModelName('testing gpt-4')).toBe('gpt-4');
+    });
+
+    test('should return "gpt-3.5-turbo" for input that does not meet any specific condition', () => {
+      expect(client.getFunctionModelName('random string')).toBe('gpt-3.5-turbo');
+      expect(client.getFunctionModelName('')).toBe('gpt-3.5-turbo');
+    });
+  });
+
+  describe('Azure OpenAI tests specific to Plugins', () => {
+    // TODO: add more tests for Azure OpenAI integration with Plugins
+    // let client;
+    // beforeEach(() => {
+    //   client = new PluginsClient('dummy_api_key');
+    // });
+
+    test('should not call getFunctionModelName when azure options are set', () => {
+      const spy = jest.spyOn(PluginsClient.prototype, 'getFunctionModelName');
+      const model = 'gpt-4-turbo';
+
+      // note, without the azure change in PR #1766, `getFunctionModelName` is called twice
+      const testClient = new PluginsClient('dummy_api_key', {
+        agentOptions: {
+          model,
+          agent: 'functions',
+        },
+        azure: defaultAzureOptions,
+      });
+
+      expect(spy).not.toHaveBeenCalled();
+      expect(testClient.agentOptions.model).toBe(model);
+
+      spy.mockRestore();
+    });
+  });
+
+  describe('sendMessage with filtered tools', () => {
+    let TestAgent;
+    const apiKey = 'fake-api-key';
+    const mockTools = [{ name: 'tool1' }, { name: 'tool2' }, { name: 'tool3' }, { name: 'tool4' }];
+
+    beforeEach(() => {
+      TestAgent = new PluginsClient(apiKey, {
+        tools: mockTools,
+        modelOptions: {
+          model: 'gpt-3.5-turbo',
+          temperature: 0,
+          max_tokens: 2,
+        },
+        agentOptions: {
+          model: 'gpt-3.5-turbo',
+        },
+      });
+
+      TestAgent.options.req = {
+        app: {
+          locals: {},
+        },
+      };
+
+      TestAgent.sendMessage = jest.fn().mockImplementation(async () => {
+        const { filteredTools = [], includedTools = [] } = TestAgent.options.req.app.locals;
+
+        if (includedTools.length > 0) {
+          const tools = TestAgent.options.tools.filter((plugin) =>
+            includedTools.includes(plugin.name),
+          );
+          TestAgent.options.tools = tools;
+        } else {
+          const tools = TestAgent.options.tools.filter(
+            (plugin) => !filteredTools.includes(plugin.name),
+          );
+          TestAgent.options.tools = tools;
+        }
+
+        return {
+          text: 'Mocked response',
+          tools: TestAgent.options.tools,
+        };
+      });
+    });
+
+    test('should filter out tools when filteredTools is provided', async () => {
+      TestAgent.options.req.app.locals.filteredTools = ['tool1', 'tool3'];
+      const response = await TestAgent.sendMessage('Test message');
+      expect(response.tools).toHaveLength(2);
+      expect(response.tools).toEqual(
+        expect.arrayContaining([
+          expect.objectContaining({ name: 'tool2' }),
+          expect.objectContaining({ name: 'tool4' }),
+        ]),
+      );
+    });
+
+    test('should only include specified tools when includedTools is provided', async () => {
+      TestAgent.options.req.app.locals.includedTools = ['tool2', 'tool4'];
+      const response = await TestAgent.sendMessage('Test message');
+      expect(response.tools).toHaveLength(2);
+      expect(response.tools).toEqual(
+        expect.arrayContaining([
+          expect.objectContaining({ name: 'tool2' }),
+          expect.objectContaining({ name: 'tool4' }),
+        ]),
+      );
+    });
+
+    test('should prioritize includedTools over filteredTools', async () => {
+      TestAgent.options.req.app.locals.filteredTools = ['tool1', 'tool3'];
+      TestAgent.options.req.app.locals.includedTools = ['tool1', 'tool2'];
+      const response = await TestAgent.sendMessage('Test message');
+      expect(response.tools).toHaveLength(2);
+      expect(response.tools).toEqual(
+        expect.arrayContaining([
+          expect.objectContaining({ name: 'tool1' }),
+          expect.objectContaining({ name: 'tool2' }),
+        ]),
+      );
+    });
+
+    test('should not modify tools when no filters are provided', async () => {
+      const response = await TestAgent.sendMessage('Test message');
+      expect(response.tools).toHaveLength(4);
+      expect(response.tools).toEqual(expect.arrayContaining(mockTools));
+    });
+  });
+});
--- a/api/app/clients/tools/dynamic/OpenAPIPlugin.js
+++ b/api/app/clients/tools/dynamic/OpenAPIPlugin.js
@@ -0,0 +1,184 @@
+require('dotenv').config();
+const fs = require('fs');
+const { z } = require('zod');
+const path = require('path');
+const yaml = require('js-yaml');
+const { createOpenAPIChain } = require('langchain/chains');
+const { DynamicStructuredTool } = require('@langchain/core/tools');
+const { ChatPromptTemplate, HumanMessagePromptTemplate } = require('@langchain/core/prompts');
+const { logger } = require('~/config');
+
+function addLinePrefix(text, prefix = '// ') {
+  return text
+    .split('\n')
+    .map((line) => prefix + line)
+    .join('\n');
+}
+
+function createPrompt(name, functions) {
+  const prefix = `// The ${name} tool has the following functions. Determine the desired or most optimal function for the user's query:`;
+  const functionDescriptions = functions
+    .map((func) => `// - ${func.name}: ${func.description}`)
+    .join('\n');
+  return `${prefix}\n${functionDescriptions}
+// You are an expert manager and scrum master. You must provide a detailed intent to better execute the function.
+// Always format as such: {{"func": "function_name", "intent": "intent and expected result"}}`;
+}
+
+const AuthBearer = z
+  .object({
+    type: z.string().includes('service_http'),
+    authorization_type: z.string().includes('bearer'),
+    verification_tokens: z.object({
+      openai: z.string(),
+    }),
+  })
+  .catch(() => false);
+
+const AuthDefinition = z
+  .object({
+    type: z.string(),
+    authorization_type: z.string(),
+    verification_tokens: z.object({
+      openai: z.string(),
+    }),
+  })
+  .catch(() => false);
+
+async function readSpecFile(filePath) {
+  try {
+    const fileContents = await fs.promises.readFile(filePath, 'utf8');
+    if (path.extname(filePath) === '.json') {
+      return JSON.parse(fileContents);
+    }
+    return yaml.load(fileContents);
+  } catch (e) {
+    logger.error('[readSpecFile] error', e);
+    return false;
+  }
+}
+
+async function getSpec(url) {
+  const RegularUrl = z
+    .string()
+    .url()
+    .catch(() => false);
+
+  if (RegularUrl.parse(url) && path.extname(url) === '.json') {
+    const response = await fetch(url);
+    return await response.json();
+  }
+
+  const ValidSpecPath = z
+    .string()
+    .url()
+    .catch(async () => {
+      const spec = path.join(__dirname, '..', '.well-known', 'openapi', url);
+      if (!fs.existsSync(spec)) {
+        return false;
+      }
+
+      return await readSpecFile(spec);
+    });
+
+  return ValidSpecPath.parse(url);
+}
+
+async function createOpenAPIPlugin({ data, llm, user, message, memory, signal }) {
+  let spec;
+  try {
+    spec = await getSpec(data.api.url);
+  } catch (error) {
+    logger.error('[createOpenAPIPlugin] getSpec error', error);
+    return null;
+  }
+
+  if (!spec) {
+    logger.warn('[createOpenAPIPlugin] No spec found');
+    return null;
+  }
+
+  const headers = {};
+  const { auth, name_for_model, description_for_model, description_for_human } = data;
+  if (auth && AuthDefinition.parse(auth)) {
+    logger.debug('[createOpenAPIPlugin] auth detected', auth);
+    const { openai } = auth.verification_tokens;
+    if (AuthBearer.parse(auth)) {
+      headers.authorization = `Bearer ${openai}`;
+      logger.debug('[createOpenAPIPlugin] added auth bearer', headers);
+    }
+  }
+
+  const chainOptions = { llm };
+
+  if (data.headers && data.headers['librechat_user_id']) {
+    logger.debug('[createOpenAPIPlugin] id detected', headers);
+    headers[data.headers['librechat_user_id']] = user;
+  }
+
+  if (Object.keys(headers).length > 0) {
+    logger.debug('[createOpenAPIPlugin] headers detected', headers);
+    chainOptions.headers = headers;
+  }
+
+  if (data.params) {
+    logger.debug('[createOpenAPIPlugin] params detected', data.params);
+    chainOptions.params = data.params;
+  }
+
+  let history = '';
+  if (memory) {
+    logger.debug('[createOpenAPIPlugin] openAPI chain: memory detected', memory);
+    const { history: chat_history } = await memory.loadMemoryVariables({});
+    history = chat_history?.length > 0 ? `\n\n## Chat History:\n${chat_history}\n` : '';
+  }
+
+  chainOptions.prompt = ChatPromptTemplate.fromMessages([
+    HumanMessagePromptTemplate.fromTemplate(
+      `# Use the provided API's to respond to this query:\n\n{query}\n\n## Instructions:\n${addLinePrefix(
+        description_for_model,
+      )}${history}`,
+    ),
+  ]);
+
+  const chain = await createOpenAPIChain(spec, chainOptions);
+
+  const { functions } = chain.chains[0].lc_kwargs.llmKwargs;
+
+  return new DynamicStructuredTool({
+    name: name_for_model,
+    description_for_model: `${addLinePrefix(description_for_human)}${createPrompt(
+      name_for_model,
+      functions,
+    )}`,
+    description: `${description_for_human}`,
+    schema: z.object({
+      func: z
+        .string()
+        .describe(
+          `The function to invoke. The functions available are: ${functions
+            .map((func) => func.name)
+            .join(', ')}`,
+        ),
+      intent: z
+        .string()
+        .describe('Describe your intent with the function and your expected result'),
+    }),
+    func: async ({ func = '', intent = '' }) => {
+      const filteredFunctions = functions.filter((f) => f.name === func);
+      chain.chains[0].lc_kwargs.llmKwargs.functions = filteredFunctions;
+      const query = `${message}${func?.length > 0 ? `\n// Intent: ${intent}` : ''}`;
+      const result = await chain.call({
+        query,
+        signal,
+      });
+      return result.response;
+    },
+  });
+}
+
+module.exports = {
+  getSpec,
+  readSpecFile,
+  createOpenAPIPlugin,
+};
--- a/api/app/clients/tools/dynamic/OpenAPIPlugin.spec.js
+++ b/api/app/clients/tools/dynamic/OpenAPIPlugin.spec.js
@@ -0,0 +1,72 @@
+const fs = require('fs');
+const { createOpenAPIPlugin, getSpec, readSpecFile } = require('./OpenAPIPlugin');
+
+global.fetch = jest.fn().mockImplementationOnce(() => {
+  return new Promise((resolve) => {
+    resolve({
+      ok: true,
+      json: () => Promise.resolve({ key: 'value' }),
+    });
+  });
+});
+jest.mock('fs', () => ({
+  promises: {
+    readFile: jest.fn(),
+  },
+  existsSync: jest.fn(),
+}));
+
+describe('readSpecFile', () => {
+  it('reads JSON file correctly', async () => {
+    fs.promises.readFile.mockResolvedValue(JSON.stringify({ test: 'value' }));
+    const result = await readSpecFile('test.json');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('reads YAML file correctly', async () => {
+    fs.promises.readFile.mockResolvedValue('test: value');
+    const result = await readSpecFile('test.yaml');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('handles error correctly', async () => {
+    fs.promises.readFile.mockRejectedValue(new Error('test error'));
+    const result = await readSpecFile('test.json');
+    expect(result).toBe(false);
+  });
+});
+
+describe('getSpec', () => {
+  it('fetches spec from url correctly', async () => {
+    const parsedJson = await getSpec('https://www.instacart.com/.well-known/ai-plugin.json');
+    const isObject = typeof parsedJson === 'object';
+    expect(isObject).toEqual(true);
+  });
+
+  it('reads spec from file correctly', async () => {
+    fs.existsSync.mockReturnValue(true);
+    fs.promises.readFile.mockResolvedValue(JSON.stringify({ test: 'value' }));
+    const result = await getSpec('test.json');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('returns false when file does not exist', async () => {
+    fs.existsSync.mockReturnValue(false);
+    const result = await getSpec('test.json');
+    expect(result).toBe(false);
+  });
+});
+
+describe('createOpenAPIPlugin', () => {
+  it('returns null when getSpec throws an error', async () => {
+    const result = await createOpenAPIPlugin({ data: { api: { url: 'invalid' } } });
+    expect(result).toBe(null);
+  });
+
+  it('returns null when no spec is found', async () => {
+    const result = await createOpenAPIPlugin({});
+    expect(result).toBe(null);
+  });
+
+  // Add more tests here for different scenarios
+});
--- a/api/app/clients/tools/index.js
+++ b/api/app/clients/tools/index.js
@@ -1,30 +1,22 @@
-const manifest = require('./manifest');
+const availableTools = require('./manifest.json');

 // Structured Tools
 const DALLE3 = require('./structured/DALLE3');
-const FluxAPI = require('./structured/FluxAPI');
-const OpenWeather = require('./structured/OpenWeather');
 const StructuredWolfram = require('./structured/Wolfram');
-const createYouTubeTools = require('./structured/YouTube');
 const StructuredACS = require('./structured/AzureAISearch');
 const StructuredSD = require('./structured/StableDiffusion');
 const GoogleSearchAPI = require('./structured/GoogleSearch');
 const TraversaalSearch = require('./structured/TraversaalSearch');
-const createOpenAIImageTools = require('./structured/OpenAIImageTools');
 const TavilySearchResults = require('./structured/TavilySearchResults');

 module.exports = {
-  ...manifest,
+  availableTools,
  // Structured Tools
  DALLE3,
-  FluxAPI,
-  OpenWeather,
  StructuredSD,
  StructuredACS,
  GoogleSearchAPI,
  TraversaalSearch,
  StructuredWolfram,
-  createYouTubeTools,
  TavilySearchResults,
-  createOpenAIImageTools,
 };
--- a/api/app/clients/tools/manifest.js
+++ b/api/app/clients/tools/manifest.js
@@ -1,20 +0,0 @@
-const availableTools = require('./manifest.json');
-
-/** @type {Record<string, TPlugin | undefined>} */
-const manifestToolMap = {};
-
-/** @type {Array<TPlugin>} */
-const toolkits = [];
-
-availableTools.forEach((tool) => {
-  manifestToolMap[tool.pluginKey] = tool;
-  if (tool.toolkit === true) {
-    toolkits.push(tool);
-  }
-});
-
-module.exports = {
-  toolkits,
-  availableTools,
-  manifestToolMap,
-};
--- a/api/app/clients/tools/manifest.json
+++ b/api/app/clients/tools/manifest.json
@@ -30,34 +30,6 @@
      }
    ]
  },
-  {
-    "name": "YouTube",
-    "pluginKey": "youtube",
-    "toolkit": true,
-    "description": "Get YouTube video information, retrieve comments, analyze transcripts and search for videos.",
-    "icon": "https://www.youtube.com/s/desktop/7449ebf7/img/favicon_144x144.png",
-    "authConfig": [
-      {
-        "authField": "YOUTUBE_API_KEY",
-        "label": "YouTube API Key",
-        "description": "Your YouTube Data API v3 key."
-      }
-    ]
-  },
-  {
-    "name": "OpenAI Image Tools",
-    "pluginKey": "image_gen_oai",
-    "toolkit": true,
-    "description": "Image Generation and Editing using OpenAI's latest state-of-the-art models",
-    "icon": "assets/image_gen_oai.png",
-    "authConfig": [
-      {
-        "authField": "IMAGE_GEN_OAI_API_KEY",
-        "label": "OpenAI Image Tools API Key",
-        "description": "Your OpenAI API Key for Image Generation and Editing"
-      }
-    ]
-  },
  {
    "name": "Wolfram",
    "pluginKey": "wolfram",
@@ -75,7 +47,7 @@
    "name": "Browser",
    "pluginKey": "web-browser",
    "description": "Scrape and summarize webpage data",
-    "icon": "assets/web-browser.svg",
+    "icon": "/assets/web-browser.svg",
    "authConfig": [
      {
        "authField": "OPENAI_API_KEY",
@@ -128,6 +100,7 @@
    "pluginKey": "calculator",
    "description": "Perform simple and complex mathematical calculations.",
    "icon": "https://i.imgur.com/RHsSG5h.png",
+    "isAuthRequired": "false",
    "authConfig": []
  },
  {
@@ -162,34 +135,7 @@
      {
        "authField": "AZURE_AI_SEARCH_API_KEY",
        "label": "Azure AI Search API Key",
-        "description": "You need to provide your API Key for Azure AI Search."
-      }
-    ]
-  },
-  {
-    "name": "OpenWeather",
-    "pluginKey": "open_weather",
-    "description": "Get weather forecasts and historical data from the OpenWeather API",
-    "icon": "assets/openweather.png",
-    "authConfig": [
-      {
-        "authField": "OPENWEATHER_API_KEY",
-        "label": "OpenWeather API Key",
-        "description": "Sign up at <a href=\"https://home.openweathermap.org/users/sign_up\" target=\"_blank\">OpenWeather</a>, then get your key at <a href=\"https://home.openweathermap.org/api_keys\" target=\"_blank\">API keys</a>."
-      }
-    ]
-  },
-  {
-    "name": "Flux",
-    "pluginKey": "flux",
-    "description": "Generate images using text with the Flux API.",
-    "icon": "https://blackforestlabs.ai/wp-content/uploads/2024/07/bfl_logo_retraced_blk.png",
-    "isAuthRequired": "true",
-    "authConfig": [
-      {
-        "authField": "FLUX_API_KEY",
-        "label": "Your Flux API Key",
-        "description": "Provide your Flux API key from your user profile."
+        "description": "You need to provideq your API Key for Azure AI Search."
      }
    ]
  }
--- a/api/app/clients/tools/structured/DALLE3.js
+++ b/api/app/clients/tools/structured/DALLE3.js
@@ -1,17 +1,14 @@
 const { z } = require('zod');
 const path = require('path');
 const OpenAI = require('openai');
-const fetch = require('node-fetch');
 const { v4: uuidv4 } = require('uuid');
-const { ProxyAgent } = require('undici');
 const { Tool } = require('@langchain/core/tools');
-const { logger } = require('@librechat/data-schemas');
-const { getImageBasename } = require('@librechat/api');
-const { FileContext, ContentTypes } = require('librechat-data-provider');
+const { HttpsProxyAgent } = require('https-proxy-agent');
+const { FileContext } = require('librechat-data-provider');
+const { getImageBasename } = require('~/server/services/Files/images');
 const extractBaseURL = require('~/utils/extractBaseURL');
+const { logger } = require('~/config');

-const displayMessage =
-  "DALL-E displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
 class DALLE3 extends Tool {
  constructor(fields = {}) {
    super();
@@ -22,8 +19,6 @@ class DALLE3 extends Tool {

    this.userId = fields.userId;
    this.fileStrategy = fields.fileStrategy;
-    /** @type {boolean} */
-    this.isAgent = fields.isAgent;
    if (fields.processFileURL) {
      /** @type {processFileURL} Necessary for output to contain all image metadata. */
      this.processFileURL = fields.processFileURL.bind(this);
@@ -46,10 +41,7 @@ class DALLE3 extends Tool {
    }

    if (process.env.PROXY) {
-      const proxyAgent = new ProxyAgent(process.env.PROXY);
-      config.fetchOptions = {
-        dispatcher: proxyAgent,
-      };
+      config.httpAgent = new HttpsProxyAgent(process.env.PROXY);
    }

    /** @type {OpenAI} */
@@ -116,16 +108,6 @@ class DALLE3 extends Tool {
    return `![generated image](${imageUrl})`;
  }

-  returnValue(value) {
-    if (this.isAgent === true && typeof value === 'string') {
-      return [value, {}];
-    } else if (this.isAgent === true && typeof value === 'object') {
-      return [displayMessage, value];
-    }
-
-    return value;
-  }
-
  async _call(data) {
    const { prompt, quality = 'standard', size = '1024x1024', style = 'vivid' } = data;
    if (!prompt) {
@@ -144,50 +126,18 @@ class DALLE3 extends Tool {
      });
    } catch (error) {
      logger.error('[DALL-E-3] Problem generating the image:', error);
-      return this
-        .returnValue(`Something went wrong when trying to generate the image. The DALL-E API may be unavailable:
-Error Message: ${error.message}`);
+      return `Something went wrong when trying to generate the image. The DALL-E API may be unavailable:
+Error Message: ${error.message}`;
    }

    if (!resp) {
-      return this.returnValue(
-        'Something went wrong when trying to generate the image. The DALL-E API may be unavailable',
-      );
+      return 'Something went wrong when trying to generate the image. The DALL-E API may be unavailable';
    }

    const theImageUrl = resp.data[0].url;

    if (!theImageUrl) {
-      return this.returnValue(
-        'No image URL returned from OpenAI API. There may be a problem with the API or your configuration.',
-      );
-    }
-
-    if (this.isAgent) {
-      let fetchOptions = {};
-      if (process.env.PROXY) {
-        const proxyAgent = new ProxyAgent(process.env.PROXY);
-        fetchOptions.dispatcher = proxyAgent;
-      }
-      const imageResponse = await fetch(theImageUrl, fetchOptions);
-      const arrayBuffer = await imageResponse.arrayBuffer();
-      const base64 = Buffer.from(arrayBuffer).toString('base64');
-      const content = [
-        {
-          type: ContentTypes.IMAGE_URL,
-          image_url: {
-            url: `data:image/png;base64,${base64}`,
-          },
-        },
-      ];
-
-      const response = [
-        {
-          type: ContentTypes.TEXT,
-          text: displayMessage,
-        },
-      ];
-      return [response, { content }];
+      return 'No image URL returned from OpenAI API. There may be a problem with the API or your configuration.';
    }

    const imageBasename = getImageBasename(theImageUrl);
@@ -207,11 +157,11 @@ Error Message: ${error.message}`);

    try {
      const result = await this.processFileURL({
-        URL: theImageUrl,
-        basePath: 'images',
-        userId: this.userId,
-        fileName: imageName,
        fileStrategy: this.fileStrategy,
+        userId: this.userId,
+        URL: theImageUrl,
+        fileName: imageName,
+        basePath: 'images',
        context: FileContext.image_generation,
      });

@@ -225,7 +175,7 @@ Error Message: ${error.message}`);
      this.result = `Failed to save the image locally. ${error.message}`;
    }

-    return this.returnValue(this.result);
+    return this.result;
  }
 }

--- a/api/app/clients/tools/structured/FluxAPI.js
+++ b/api/app/clients/tools/structured/FluxAPI.js
@@ -1,554 +0,0 @@
-const { z } = require('zod');
-const axios = require('axios');
-const fetch = require('node-fetch');
-const { v4: uuidv4 } = require('uuid');
-const { Tool } = require('@langchain/core/tools');
-const { HttpsProxyAgent } = require('https-proxy-agent');
-const { FileContext, ContentTypes } = require('librechat-data-provider');
-const { logger } = require('~/config');
-
-const displayMessage =
-  'Flux displayed an image. All generated images are already plainly visible, so don\'t repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.';
-
-/**
- * FluxAPI - A tool for generating high-quality images from text prompts using the Flux API.
- * Each call generates one image. If multiple images are needed, make multiple consecutive calls with the same or varied prompts.
- */
-class FluxAPI extends Tool {
-  // Pricing constants in USD per image
-  static PRICING = {
-    FLUX_PRO_1_1_ULTRA: -0.06, // /v1/flux-pro-1.1-ultra
-    FLUX_PRO_1_1: -0.04, // /v1/flux-pro-1.1
-    FLUX_PRO: -0.05, // /v1/flux-pro
-    FLUX_DEV: -0.025, // /v1/flux-dev
-    FLUX_PRO_FINETUNED: -0.06, // /v1/flux-pro-finetuned
-    FLUX_PRO_1_1_ULTRA_FINETUNED: -0.07, // /v1/flux-pro-1.1-ultra-finetuned
-  };
-
-  constructor(fields = {}) {
-    super();
-
-    /** @type {boolean} Used to initialize the Tool without necessary variables. */
-    this.override = fields.override ?? false;
-
-    this.userId = fields.userId;
-    this.fileStrategy = fields.fileStrategy;
-
-    /** @type {boolean} **/
-    this.isAgent = fields.isAgent;
-    this.returnMetadata = fields.returnMetadata ?? false;
-
-    if (fields.processFileURL) {
-      /** @type {processFileURL} Necessary for output to contain all image metadata. */
-      this.processFileURL = fields.processFileURL.bind(this);
-    }
-
-    this.apiKey = fields.FLUX_API_KEY || this.getApiKey();
-
-    this.name = 'flux';
-    this.description =
-      'Use Flux to generate images from text descriptions. This tool can generate images and list available finetunes. Each generate call creates one image. For multiple images, make multiple consecutive calls.';
-
-    this.description_for_model = `// Transform any image description into a detailed, high-quality prompt. Never submit a prompt under 3 sentences. Follow these core rules:
-    // 1. ALWAYS enhance basic prompts into 5-10 detailed sentences (e.g., "a cat" becomes: "A close-up photo of a sleek Siamese cat with piercing blue eyes. The cat sits elegantly on a vintage leather armchair, its tail curled gracefully around its paws. Warm afternoon sunlight streams through a nearby window, casting gentle shadows across its face and highlighting the subtle variations in its cream and chocolate-point fur. The background is softly blurred, creating a shallow depth of field that draws attention to the cat's expressive features. The overall composition has a peaceful, contemplative mood with a professional photography style.")
-    // 2. Each prompt MUST be 3-6 descriptive sentences minimum, focusing on visual elements: lighting, composition, mood, and style
-    // Use action: 'list_finetunes' to see available custom models. When using finetunes, use endpoint: '/v1/flux-pro-finetuned' (default) or '/v1/flux-pro-1.1-ultra-finetuned' for higher quality and aspect ratio.`;
-
-    // Add base URL from environment variable with fallback
-    this.baseUrl = process.env.FLUX_API_BASE_URL || 'https://api.us1.bfl.ai';
-
-    // Define the schema for structured input
-    this.schema = z.object({
-      action: z
-        .enum(['generate', 'list_finetunes', 'generate_finetuned'])
-        .default('generate')
-        .describe(
-          'Action to perform: "generate" for image generation, "generate_finetuned" for finetuned model generation, "list_finetunes" to get available custom models',
-        ),
-      prompt: z
-        .string()
-        .optional()
-        .describe(
-          'Text prompt for image generation. Required when action is "generate". Not used for list_finetunes.',
-        ),
-      width: z
-        .number()
-        .optional()
-        .describe(
-          'Width of the generated image in pixels. Must be a multiple of 32. Default is 1024.',
-        ),
-      height: z
-        .number()
-        .optional()
-        .describe(
-          'Height of the generated image in pixels. Must be a multiple of 32. Default is 768.',
-        ),
-      prompt_upsampling: z
-        .boolean()
-        .optional()
-        .default(false)
-        .describe('Whether to perform upsampling on the prompt.'),
-      steps: z
-        .number()
-        .int()
-        .optional()
-        .describe('Number of steps to run the model for, a number from 1 to 50. Default is 40.'),
-      seed: z.number().optional().describe('Optional seed for reproducibility.'),
-      safety_tolerance: z
-        .number()
-        .optional()
-        .default(6)
-        .describe(
-          'Tolerance level for input and output moderation. Between 0 and 6, 0 being most strict, 6 being least strict.',
-        ),
-      endpoint: z
-        .enum([
-          '/v1/flux-pro-1.1',
-          '/v1/flux-pro',
-          '/v1/flux-dev',
-          '/v1/flux-pro-1.1-ultra',
-          '/v1/flux-pro-finetuned',
-          '/v1/flux-pro-1.1-ultra-finetuned',
-        ])
-        .optional()
-        .default('/v1/flux-pro-1.1')
-        .describe('Endpoint to use for image generation.'),
-      raw: z
-        .boolean()
-        .optional()
-        .default(false)
-        .describe(
-          'Generate less processed, more natural-looking images. Only works for /v1/flux-pro-1.1-ultra.',
-        ),
-      finetune_id: z.string().optional().describe('ID of the finetuned model to use'),
-      finetune_strength: z
-        .number()
-        .optional()
-        .default(1.1)
-        .describe('Strength of the finetuning effect (typically between 0.1 and 1.2)'),
-      guidance: z.number().optional().default(2.5).describe('Guidance scale for finetuned models'),
-      aspect_ratio: z
-        .string()
-        .optional()
-        .default('16:9')
-        .describe('Aspect ratio for ultra models (e.g., "16:9")'),
-    });
-  }
-
-  getAxiosConfig() {
-    const config = {};
-    if (process.env.PROXY) {
-      config.httpsAgent = new HttpsProxyAgent(process.env.PROXY);
-    }
-    return config;
-  }
-
-  /** @param {Object|string} value */
-  getDetails(value) {
-    if (typeof value === 'string') {
-      return value;
-    }
-    return JSON.stringify(value, null, 2);
-  }
-
-  getApiKey() {
-    const apiKey = process.env.FLUX_API_KEY || '';
-    if (!apiKey && !this.override) {
-      throw new Error('Missing FLUX_API_KEY environment variable.');
-    }
-    return apiKey;
-  }
-
-  wrapInMarkdown(imageUrl) {
-    const serverDomain = process.env.DOMAIN_SERVER || 'http://localhost:3080';
-    return `![generated image](${serverDomain}${imageUrl})`;
-  }
-
-  returnValue(value) {
-    if (this.isAgent === true && typeof value === 'string') {
-      return [value, {}];
-    } else if (this.isAgent === true && typeof value === 'object') {
-      if (Array.isArray(value)) {
-        return value;
-      }
-      return [displayMessage, value];
-    }
-    return value;
-  }
-
-  async _call(data) {
-    const { action = 'generate', ...imageData } = data;
-
-    // Use provided API key for this request if available, otherwise use default
-    const requestApiKey = this.apiKey || this.getApiKey();
-
-    // Handle list_finetunes action
-    if (action === 'list_finetunes') {
-      return this.getMyFinetunes(requestApiKey);
-    }
-
-    // Handle finetuned generation
-    if (action === 'generate_finetuned') {
-      return this.generateFinetunedImage(imageData, requestApiKey);
-    }
-
-    // For generate action, ensure prompt is provided
-    if (!imageData.prompt) {
-      throw new Error('Missing required field: prompt');
-    }
-
-    let payload = {
-      prompt: imageData.prompt,
-      prompt_upsampling: imageData.prompt_upsampling || false,
-      safety_tolerance: imageData.safety_tolerance || 6,
-      output_format: imageData.output_format || 'png',
-    };
-
-    // Add optional parameters if provided
-    if (imageData.width) {
-      payload.width = imageData.width;
-    }
-    if (imageData.height) {
-      payload.height = imageData.height;
-    }
-    if (imageData.steps) {
-      payload.steps = imageData.steps;
-    }
-    if (imageData.seed !== undefined) {
-      payload.seed = imageData.seed;
-    }
-    if (imageData.raw) {
-      payload.raw = imageData.raw;
-    }
-
-    const generateUrl = `${this.baseUrl}${imageData.endpoint || '/v1/flux-pro'}`;
-    const resultUrl = `${this.baseUrl}/v1/get_result`;
-
-    logger.debug('[FluxAPI] Generating image with payload:', payload);
-    logger.debug('[FluxAPI] Using endpoint:', generateUrl);
-
-    let taskResponse;
-    try {
-      taskResponse = await axios.post(generateUrl, payload, {
-        headers: {
-          'x-key': requestApiKey,
-          'Content-Type': 'application/json',
-          Accept: 'application/json',
-        },
-        ...this.getAxiosConfig(),
-      });
-    } catch (error) {
-      const details = this.getDetails(error?.response?.data || error.message);
-      logger.error('[FluxAPI] Error while submitting task:', details);
-
-      return this.returnValue(
-        `Something went wrong when trying to generate the image. The Flux API may be unavailable:
-        Error Message: ${details}`,
-      );
-    }
-
-    const taskId = taskResponse.data.id;
-
-    // Polling for the result
-    let status = 'Pending';
-    let resultData = null;
-    while (status !== 'Ready' && status !== 'Error') {
-      try {
-        // Wait 2 seconds between polls
-        await new Promise((resolve) => setTimeout(resolve, 2000));
-        const resultResponse = await axios.get(resultUrl, {
-          headers: {
-            'x-key': requestApiKey,
-            Accept: 'application/json',
-          },
-          params: { id: taskId },
-          ...this.getAxiosConfig(),
-        });
-        status = resultResponse.data.status;
-
-        if (status === 'Ready') {
-          resultData = resultResponse.data.result;
-          break;
-        } else if (status === 'Error') {
-          logger.error('[FluxAPI] Error in task:', resultResponse.data);
-          return this.returnValue('An error occurred during image generation.');
-        }
-      } catch (error) {
-        const details = this.getDetails(error?.response?.data || error.message);
-        logger.error('[FluxAPI] Error while getting result:', details);
-        return this.returnValue('An error occurred while retrieving the image.');
-      }
-    }
-
-    // If no result data
-    if (!resultData || !resultData.sample) {
-      logger.error('[FluxAPI] No image data received from API. Response:', resultData);
-      return this.returnValue('No image data received from Flux API.');
-    }
-
-    // Try saving the image locally
-    const imageUrl = resultData.sample;
-    const imageName = `img-${uuidv4()}.png`;
-
-    if (this.isAgent) {
-      try {
-        // Fetch the image and convert to base64
-        const fetchOptions = {};
-        if (process.env.PROXY) {
-          fetchOptions.agent = new HttpsProxyAgent(process.env.PROXY);
-        }
-        const imageResponse = await fetch(imageUrl, fetchOptions);
-        const arrayBuffer = await imageResponse.arrayBuffer();
-        const base64 = Buffer.from(arrayBuffer).toString('base64');
-        const content = [
-          {
-            type: ContentTypes.IMAGE_URL,
-            image_url: {
-              url: `data:image/png;base64,${base64}`,
-            },
-          },
-        ];
-
-        const response = [
-          {
-            type: ContentTypes.TEXT,
-            text: displayMessage,
-          },
-        ];
-        return [response, { content }];
-      } catch (error) {
-        logger.error('Error processing image for agent:', error);
-        return this.returnValue(`Failed to process the image. ${error.message}`);
-      }
-    }
-
-    try {
-      logger.debug('[FluxAPI] Saving image:', imageUrl);
-      const result = await this.processFileURL({
-        fileStrategy: this.fileStrategy,
-        userId: this.userId,
-        URL: imageUrl,
-        fileName: imageName,
-        basePath: 'images',
-        context: FileContext.image_generation,
-      });
-
-      logger.debug('[FluxAPI] Image saved to path:', result.filepath);
-
-      // Calculate cost based on endpoint
-      /**
-       * TODO: Cost handling
-      const endpoint = imageData.endpoint || '/v1/flux-pro';
-      const endpointKey = Object.entries(FluxAPI.PRICING).find(([key, _]) =>
-        endpoint.includes(key.toLowerCase().replace(/_/g, '-')),
-      )?.[0];
-      const cost = FluxAPI.PRICING[endpointKey] || 0;
-       */
-      this.result = this.returnMetadata ? result : this.wrapInMarkdown(result.filepath);
-      return this.returnValue(this.result);
-    } catch (error) {
-      const details = this.getDetails(error?.message ?? 'No additional error details.');
-      logger.error('Error while saving the image:', details);
-      return this.returnValue(`Failed to save the image locally. ${details}`);
-    }
-  }
-
-  async getMyFinetunes(apiKey = null) {
-    const finetunesUrl = `${this.baseUrl}/v1/my_finetunes`;
-    const detailsUrl = `${this.baseUrl}/v1/finetune_details`;
-
-    try {
-      const headers = {
-        'x-key': apiKey || this.getApiKey(),
-        'Content-Type': 'application/json',
-        Accept: 'application/json',
-      };
-
-      // Get list of finetunes
-      const response = await axios.get(finetunesUrl, {
-        headers,
-        ...this.getAxiosConfig(),
-      });
-      const finetunes = response.data.finetunes;
-
-      // Fetch details for each finetune
-      const finetuneDetails = await Promise.all(
-        finetunes.map(async (finetuneId) => {
-          try {
-            const detailResponse = await axios.get(`${detailsUrl}?finetune_id=${finetuneId}`, {
-              headers,
-              ...this.getAxiosConfig(),
-            });
-            return {
-              id: finetuneId,
-              ...detailResponse.data,
-            };
-          } catch (error) {
-            logger.error(`[FluxAPI] Error fetching details for finetune ${finetuneId}:`, error);
-            return {
-              id: finetuneId,
-              error: 'Failed to fetch details',
-            };
-          }
-        }),
-      );
-
-      if (this.isAgent) {
-        const formattedDetails = JSON.stringify(finetuneDetails, null, 2);
-        return [`Here are the available finetunes:\n${formattedDetails}`, null];
-      }
-      return JSON.stringify(finetuneDetails);
-    } catch (error) {
-      const details = this.getDetails(error?.response?.data || error.message);
-      logger.error('[FluxAPI] Error while getting finetunes:', details);
-      const errorMsg = `Failed to get finetunes: ${details}`;
-      return this.isAgent ? this.returnValue([errorMsg, {}]) : new Error(errorMsg);
-    }
-  }
-
-  async generateFinetunedImage(imageData, requestApiKey) {
-    if (!imageData.prompt) {
-      throw new Error('Missing required field: prompt');
-    }
-
-    if (!imageData.finetune_id) {
-      throw new Error(
-        'Missing required field: finetune_id for finetuned generation. Please supply a finetune_id!',
-      );
-    }
-
-    // Validate endpoint is appropriate for finetuned generation
-    const validFinetunedEndpoints = ['/v1/flux-pro-finetuned', '/v1/flux-pro-1.1-ultra-finetuned'];
-    const endpoint = imageData.endpoint || '/v1/flux-pro-finetuned';
-
-    if (!validFinetunedEndpoints.includes(endpoint)) {
-      throw new Error(
-        `Invalid endpoint for finetuned generation. Must be one of: ${validFinetunedEndpoints.join(', ')}`,
-      );
-    }
-
-    let payload = {
-      prompt: imageData.prompt,
-      prompt_upsampling: imageData.prompt_upsampling || false,
-      safety_tolerance: imageData.safety_tolerance || 6,
-      output_format: imageData.output_format || 'png',
-      finetune_id: imageData.finetune_id,
-      finetune_strength: imageData.finetune_strength || 1.0,
-      guidance: imageData.guidance || 2.5,
-    };
-
-    // Add optional parameters if provided
-    if (imageData.width) {
-      payload.width = imageData.width;
-    }
-    if (imageData.height) {
-      payload.height = imageData.height;
-    }
-    if (imageData.steps) {
-      payload.steps = imageData.steps;
-    }
-    if (imageData.seed !== undefined) {
-      payload.seed = imageData.seed;
-    }
-    if (imageData.raw) {
-      payload.raw = imageData.raw;
-    }
-
-    const generateUrl = `${this.baseUrl}${endpoint}`;
-    const resultUrl = `${this.baseUrl}/v1/get_result`;
-
-    logger.debug('[FluxAPI] Generating finetuned image with payload:', payload);
-    logger.debug('[FluxAPI] Using endpoint:', generateUrl);
-
-    let taskResponse;
-    try {
-      taskResponse = await axios.post(generateUrl, payload, {
-        headers: {
-          'x-key': requestApiKey,
-          'Content-Type': 'application/json',
-          Accept: 'application/json',
-        },
-        ...this.getAxiosConfig(),
-      });
-    } catch (error) {
-      const details = this.getDetails(error?.response?.data || error.message);
-      logger.error('[FluxAPI] Error while submitting finetuned task:', details);
-      return this.returnValue(
-        `Something went wrong when trying to generate the finetuned image. The Flux API may be unavailable:
-        Error Message: ${details}`,
-      );
-    }
-
-    const taskId = taskResponse.data.id;
-
-    // Polling for the result
-    let status = 'Pending';
-    let resultData = null;
-    while (status !== 'Ready' && status !== 'Error') {
-      try {
-        // Wait 2 seconds between polls
-        await new Promise((resolve) => setTimeout(resolve, 2000));
-        const resultResponse = await axios.get(resultUrl, {
-          headers: {
-            'x-key': requestApiKey,
-            Accept: 'application/json',
-          },
-          params: { id: taskId },
-          ...this.getAxiosConfig(),
-        });
-        status = resultResponse.data.status;
-
-        if (status === 'Ready') {
-          resultData = resultResponse.data.result;
-          break;
-        } else if (status === 'Error') {
-          logger.error('[FluxAPI] Error in finetuned task:', resultResponse.data);
-          return this.returnValue('An error occurred during finetuned image generation.');
-        }
-      } catch (error) {
-        const details = this.getDetails(error?.response?.data || error.message);
-        logger.error('[FluxAPI] Error while getting finetuned result:', details);
-        return this.returnValue('An error occurred while retrieving the finetuned image.');
-      }
-    }
-
-    // If no result data
-    if (!resultData || !resultData.sample) {
-      logger.error('[FluxAPI] No image data received from API. Response:', resultData);
-      return this.returnValue('No image data received from Flux API.');
-    }
-
-    // Try saving the image locally
-    const imageUrl = resultData.sample;
-    const imageName = `img-${uuidv4()}.png`;
-
-    try {
-      logger.debug('[FluxAPI] Saving finetuned image:', imageUrl);
-      const result = await this.processFileURL({
-        fileStrategy: this.fileStrategy,
-        userId: this.userId,
-        URL: imageUrl,
-        fileName: imageName,
-        basePath: 'images',
-        context: FileContext.image_generation,
-      });
-
-      logger.debug('[FluxAPI] Finetuned image saved to path:', result.filepath);
-
-      // Calculate cost based on endpoint
-      const endpointKey = endpoint.includes('ultra')
-        ? 'FLUX_PRO_1_1_ULTRA_FINETUNED'
-        : 'FLUX_PRO_FINETUNED';
-      const cost = FluxAPI.PRICING[endpointKey] || 0;
-      // Return the result based on returnMetadata flag
-      this.result = this.returnMetadata ? result : this.wrapInMarkdown(result.filepath);
-      return this.returnValue(this.result);
-    } catch (error) {
-      const details = this.getDetails(error?.message ?? 'No additional error details.');
-      logger.error('Error while saving the finetuned image:', details);
-      return this.returnValue(`Failed to save the finetuned image locally. ${details}`);
-    }
-  }
-}
-
-module.exports = FluxAPI;
--- a/api/app/clients/tools/structured/OpenAIImageTools.js
+++ b/api/app/clients/tools/structured/OpenAIImageTools.js
@@ -1,420 +0,0 @@
-const axios = require('axios');
-const { v4 } = require('uuid');
-const OpenAI = require('openai');
-const FormData = require('form-data');
-const { ProxyAgent } = require('undici');
-const { tool } = require('@langchain/core/tools');
-const { logger } = require('@librechat/data-schemas');
-const { logAxiosError, oaiToolkit } = require('@librechat/api');
-const { ContentTypes, EImageOutputType } = require('librechat-data-provider');
-const { getStrategyFunctions } = require('~/server/services/Files/strategies');
-const extractBaseURL = require('~/utils/extractBaseURL');
-const { getFiles } = require('~/models/File');
-
-const displayMessage =
-  "The tool displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
-
-/**
- * Replaces unwanted characters from the input string
- * @param {string} inputString - The input string to process
- * @returns {string} - The processed string
- */
-function replaceUnwantedChars(inputString) {
-  return inputString
-    .replace(/\r\n|\r|\n/g, ' ')
-    .replace(/"/g, '')
-    .trim();
-}
-
-function returnValue(value) {
-  if (typeof value === 'string') {
-    return [value, {}];
-  } else if (typeof value === 'object') {
-    if (Array.isArray(value)) {
-      return value;
-    }
-    return [displayMessage, value];
-  }
-  return value;
-}
-
-function createAbortHandler() {
-  return function () {
-    logger.debug('[ImageGenOAI] Image generation aborted');
-  };
-}
-
-/**
- * Creates OpenAI Image tools (generation and editing)
- * @param {Object} fields - Configuration fields
- * @param {ServerRequest} fields.req - Whether the tool is being used in an agent context
- * @param {boolean} fields.isAgent - Whether the tool is being used in an agent context
- * @param {string} fields.IMAGE_GEN_OAI_API_KEY - The OpenAI API key
- * @param {boolean} [fields.override] - Whether to override the API key check, necessary for app initialization
- * @param {MongoFile[]} [fields.imageFiles] - The images to be used for editing
- * @param {string} [fields.imageOutputType] - The image output type configuration
- * @param {string} [fields.fileStrategy] - The file storage strategy
- * @returns {Array<ReturnType<tool>>} - Array of image tools
- */
-function createOpenAIImageTools(fields = {}) {
-  /** @type {boolean} Used to initialize the Tool without necessary variables. */
-  const override = fields.override ?? false;
-  /** @type {boolean} */
-  if (!override && !fields.isAgent) {
-    throw new Error('This tool is only available for agents.');
-  }
-  const { req } = fields;
-  const imageOutputType = fields.imageOutputType || EImageOutputType.PNG;
-  const appFileStrategy = fields.fileStrategy;
-
-  const getApiKey = () => {
-    const apiKey = process.env.IMAGE_GEN_OAI_API_KEY ?? '';
-    if (!apiKey && !override) {
-      throw new Error('Missing IMAGE_GEN_OAI_API_KEY environment variable.');
-    }
-    return apiKey;
-  };
-
-  let apiKey = fields.IMAGE_GEN_OAI_API_KEY ?? getApiKey();
-  const closureConfig = { apiKey };
-
-  let baseURL = 'https://api.openai.com/v1/';
-  if (!override && process.env.IMAGE_GEN_OAI_BASEURL) {
-    baseURL = extractBaseURL(process.env.IMAGE_GEN_OAI_BASEURL);
-    closureConfig.baseURL = baseURL;
-  }
-
-  // Note: Azure may not yet support the latest image generation models
-  if (
-    !override &&
-    process.env.IMAGE_GEN_OAI_AZURE_API_VERSION &&
-    process.env.IMAGE_GEN_OAI_BASEURL
-  ) {
-    baseURL = process.env.IMAGE_GEN_OAI_BASEURL;
-    closureConfig.baseURL = baseURL;
-    closureConfig.defaultQuery = { 'api-version': process.env.IMAGE_GEN_OAI_AZURE_API_VERSION };
-    closureConfig.defaultHeaders = {
-      'api-key': process.env.IMAGE_GEN_OAI_API_KEY,
-      'Content-Type': 'application/json',
-    };
-    closureConfig.apiKey = process.env.IMAGE_GEN_OAI_API_KEY;
-  }
-
-  const imageFiles = fields.imageFiles ?? [];
-
-  /**
-   * Image Generation Tool
-   */
-  const imageGenTool = tool(
-    async (
-      {
-        prompt,
-        background = 'auto',
-        n = 1,
-        output_compression = 100,
-        quality = 'auto',
-        size = 'auto',
-      },
-      runnableConfig,
-    ) => {
-      if (!prompt) {
-        throw new Error('Missing required field: prompt');
-      }
-      const clientConfig = { ...closureConfig };
-      if (process.env.PROXY) {
-        const proxyAgent = new ProxyAgent(process.env.PROXY);
-        clientConfig.fetchOptions = {
-          dispatcher: proxyAgent,
-        };
-      }
-
-      /** @type {OpenAI} */
-      const openai = new OpenAI(clientConfig);
-      let output_format = imageOutputType;
-      if (
-        background === 'transparent' &&
-        output_format !== EImageOutputType.PNG &&
-        output_format !== EImageOutputType.WEBP
-      ) {
-        logger.warn(
-          '[ImageGenOAI] Transparent background requires PNG or WebP format, defaulting to PNG',
-        );
-        output_format = EImageOutputType.PNG;
-      }
-
-      let resp;
-      /** @type {AbortSignal} */
-      let derivedSignal = null;
-      /** @type {() => void} */
-      let abortHandler = null;
-
-      try {
-        if (runnableConfig?.signal) {
-          derivedSignal = AbortSignal.any([runnableConfig.signal]);
-          abortHandler = createAbortHandler();
-          derivedSignal.addEventListener('abort', abortHandler, { once: true });
-        }
-
-        resp = await openai.images.generate(
-          {
-            model: 'gpt-image-1',
-            prompt: replaceUnwantedChars(prompt),
-            n: Math.min(Math.max(1, n), 10),
-            background,
-            output_format,
-            output_compression:
-              output_format === EImageOutputType.WEBP || output_format === EImageOutputType.JPEG
-                ? output_compression
-                : undefined,
-            quality,
-            size,
-          },
-          {
-            signal: derivedSignal,
-          },
-        );
-      } catch (error) {
-        const message = '[image_gen_oai] Problem generating the image:';
-        logAxiosError({ error, message });
-        return returnValue(`Something went wrong when trying to generate the image. The OpenAI API may be unavailable:
-Error Message: ${error.message}`);
-      } finally {
-        if (abortHandler && derivedSignal) {
-          derivedSignal.removeEventListener('abort', abortHandler);
-        }
-      }
-
-      if (!resp) {
-        return returnValue(
-          'Something went wrong when trying to generate the image. The OpenAI API may be unavailable',
-        );
-      }
-
-      // For gpt-image-1, the response contains base64-encoded images
-      // TODO: handle cost in `resp.usage`
-      const base64Image = resp.data[0].b64_json;
-
-      if (!base64Image) {
-        return returnValue(
-          'No image data returned from OpenAI API. There may be a problem with the API or your configuration.',
-        );
-      }
-
-      const content = [
-        {
-          type: ContentTypes.IMAGE_URL,
-          image_url: {
-            url: `data:image/${output_format};base64,${base64Image}`,
-          },
-        },
-      ];
-
-      const file_ids = [v4()];
-      const response = [
-        {
-          type: ContentTypes.TEXT,
-          text: displayMessage + `\n\ngenerated_image_id: "${file_ids[0]}"`,
-        },
-      ];
-      return [response, { content, file_ids }];
-    },
-    oaiToolkit.image_gen_oai,
-  );
-
-  /**
-   * Image Editing Tool
-   */
-  const imageEditTool = tool(
-    async ({ prompt, image_ids, quality = 'auto', size = 'auto' }, runnableConfig) => {
-      if (!prompt) {
-        throw new Error('Missing required field: prompt');
-      }
-
-      const clientConfig = { ...closureConfig };
-      if (process.env.PROXY) {
-        const proxyAgent = new ProxyAgent(process.env.PROXY);
-        clientConfig.fetchOptions = {
-          dispatcher: proxyAgent,
-        };
-      }
-
-      const formData = new FormData();
-      formData.append('model', 'gpt-image-1');
-      formData.append('prompt', replaceUnwantedChars(prompt));
-      // TODO: `mask` support
-      // TODO: more than 1 image support
-      // formData.append('n', n.toString());
-      formData.append('quality', quality);
-      formData.append('size', size);
-
-      /** @type {Record<FileSources, undefined | NodeStreamDownloader<File>>} */
-      const streamMethods = {};
-
-      const requestFilesMap = Object.fromEntries(imageFiles.map((f) => [f.file_id, { ...f }]));
-
-      const orderedFiles = new Array(image_ids.length);
-      const idsToFetch = [];
-      const indexOfMissing = Object.create(null);
-
-      for (let i = 0; i < image_ids.length; i++) {
-        const id = image_ids[i];
-        const file = requestFilesMap[id];
-
-        if (file) {
-          orderedFiles[i] = file;
-        } else {
-          idsToFetch.push(id);
-          indexOfMissing[id] = i;
-        }
-      }
-
-      if (idsToFetch.length) {
-        const fetchedFiles = await getFiles(
-          {
-            user: req.user.id,
-            file_id: { $in: idsToFetch },
-            height: { $exists: true },
-            width: { $exists: true },
-          },
-          {},
-          {},
-        );
-
-        for (const file of fetchedFiles) {
-          requestFilesMap[file.file_id] = file;
-          orderedFiles[indexOfMissing[file.file_id]] = file;
-        }
-      }
-      for (const imageFile of orderedFiles) {
-        if (!imageFile) {
-          continue;
-        }
-        /** @type {NodeStream<File>} */
-        let stream;
-        /** @type {NodeStreamDownloader<File>} */
-        let getDownloadStream;
-        const source = imageFile.source || appFileStrategy;
-        if (!source) {
-          throw new Error('No source found for image file');
-        }
-        if (streamMethods[source]) {
-          getDownloadStream = streamMethods[source];
-        } else {
-          ({ getDownloadStream } = getStrategyFunctions(source));
-          streamMethods[source] = getDownloadStream;
-        }
-        if (!getDownloadStream) {
-          throw new Error(`No download stream method found for source: ${source}`);
-        }
-        stream = await getDownloadStream(req, imageFile.filepath);
-        if (!stream) {
-          throw new Error('Failed to get download stream for image file');
-        }
-        formData.append('image[]', stream, {
-          filename: imageFile.filename,
-          contentType: imageFile.type,
-        });
-      }
-
-      /** @type {import('axios').RawAxiosHeaders} */
-      let headers = {
-        ...formData.getHeaders(),
-      };
-
-      if (process.env.IMAGE_GEN_OAI_AZURE_API_VERSION && process.env.IMAGE_GEN_OAI_BASEURL) {
-        headers['api-key'] = apiKey;
-      } else {
-        headers['Authorization'] = `Bearer ${apiKey}`;
-      }
-
-      /** @type {AbortSignal} */
-      let derivedSignal = null;
-      /** @type {() => void} */
-      let abortHandler = null;
-
-      try {
-        if (runnableConfig?.signal) {
-          derivedSignal = AbortSignal.any([runnableConfig.signal]);
-          abortHandler = createAbortHandler();
-          derivedSignal.addEventListener('abort', abortHandler, { once: true });
-        }
-
-        /** @type {import('axios').AxiosRequestConfig} */
-        const axiosConfig = {
-          headers,
-          ...clientConfig,
-          signal: derivedSignal,
-          baseURL,
-        };
-
-        if (process.env.PROXY) {
-          try {
-            const url = new URL(process.env.PROXY);
-            axiosConfig.proxy = {
-              host: url.hostname.replace(/^\[|\]$/g, ''),
-              port: url.port ? parseInt(url.port, 10) : undefined,
-              protocol: url.protocol.replace(':', ''),
-            };
-          } catch (error) {
-            logger.error('Error parsing proxy URL:', error);
-          }
-        }
-
-        if (process.env.IMAGE_GEN_OAI_AZURE_API_VERSION && process.env.IMAGE_GEN_OAI_BASEURL) {
-          axiosConfig.params = {
-            'api-version': process.env.IMAGE_GEN_OAI_AZURE_API_VERSION,
-            ...axiosConfig.params,
-          };
-        }
-        const response = await axios.post('/images/edits', formData, axiosConfig);
-
-        if (!response.data || !response.data.data || !response.data.data.length) {
-          return returnValue(
-            'No image data returned from OpenAI API. There may be a problem with the API or your configuration.',
-          );
-        }
-
-        const base64Image = response.data.data[0].b64_json;
-        if (!base64Image) {
-          return returnValue(
-            'No image data returned from OpenAI API. There may be a problem with the API or your configuration.',
-          );
-        }
-
-        const content = [
-          {
-            type: ContentTypes.IMAGE_URL,
-            image_url: {
-              url: `data:image/${imageOutputType};base64,${base64Image}`,
-            },
-          },
-        ];
-
-        const file_ids = [v4()];
-        const textResponse = [
-          {
-            type: ContentTypes.TEXT,
-            text:
-              displayMessage +
-              `\n\ngenerated_image_id: "${file_ids[0]}"\nreferenced_image_ids: ["${image_ids.join('", "')}"]`,
-          },
-        ];
-        return [textResponse, { content, file_ids }];
-      } catch (error) {
-        const message = '[image_edit_oai] Problem editing the image:';
-        logAxiosError({ error, message });
-        return returnValue(`Something went wrong when trying to edit the image. The OpenAI API may be unavailable:
-Error Message: ${error.message || 'Unknown error'}`);
-      } finally {
-        if (abortHandler && derivedSignal) {
-          derivedSignal.removeEventListener('abort', abortHandler);
-        }
-      }
-    },
-    oaiToolkit.image_edit_oai,
-  );
-
-  return [imageGenTool, imageEditTool];
-}
-
-module.exports = createOpenAIImageTools;
--- a/api/app/clients/tools/structured/OpenWeather.js
+++ b/api/app/clients/tools/structured/OpenWeather.js
@@ -1,317 +0,0 @@
-const { Tool } = require('@langchain/core/tools');
-const { z } = require('zod');
-const { getEnvironmentVariable } = require('@langchain/core/utils/env');
-const fetch = require('node-fetch');
-
-/**
- * Map user-friendly units to OpenWeather units.
- * Defaults to Celsius if not specified.
- */
-function mapUnitsToOpenWeather(unit) {
-  if (!unit) {
-    return 'metric';
-  } // Default to Celsius
-  switch (unit) {
-    case 'Celsius':
-      return 'metric';
-    case 'Kelvin':
-      return 'standard';
-    case 'Fahrenheit':
-      return 'imperial';
-    default:
-      return 'metric'; // fallback
-  }
-}
-
-/**
- * Recursively round temperature fields in the API response.
- */
-function roundTemperatures(obj) {
-  const tempKeys = new Set([
-    'temp',
-    'feels_like',
-    'dew_point',
-    'day',
-    'min',
-    'max',
-    'night',
-    'eve',
-    'morn',
-    'afternoon',
-    'morning',
-    'evening',
-  ]);
-
-  if (Array.isArray(obj)) {
-    return obj.map((item) => roundTemperatures(item));
-  } else if (obj && typeof obj === 'object') {
-    for (const key of Object.keys(obj)) {
-      const value = obj[key];
-      if (value && typeof value === 'object') {
-        obj[key] = roundTemperatures(value);
-      } else if (typeof value === 'number' && tempKeys.has(key)) {
-        obj[key] = Math.round(value);
-      }
-    }
-  }
-  return obj;
-}
-
-class OpenWeather extends Tool {
-  name = 'open_weather';
-  description =
-    'Provides weather data from OpenWeather One Call API 3.0. ' +
-    'Actions: help, current_forecast, timestamp, daily_aggregation, overview. ' +
-    'If lat/lon not provided, specify "city" for geocoding. ' +
-    'Units: "Celsius", "Kelvin", or "Fahrenheit" (default: Celsius). ' +
-    'For timestamp action, use "date" in YYYY-MM-DD format.';
-
-  schema = z.object({
-    action: z.enum(['help', 'current_forecast', 'timestamp', 'daily_aggregation', 'overview']),
-    city: z.string().optional(),
-    lat: z.number().optional(),
-    lon: z.number().optional(),
-    exclude: z.string().optional(),
-    units: z.enum(['Celsius', 'Kelvin', 'Fahrenheit']).optional(),
-    lang: z.string().optional(),
-    date: z.string().optional(), // For timestamp and daily_aggregation
-    tz: z.string().optional(),
-  });
-
-  constructor(fields = {}) {
-    super();
-    this.envVar = 'OPENWEATHER_API_KEY';
-    this.override = fields.override ?? false;
-    this.apiKey = fields[this.envVar] ?? this.getApiKey();
-  }
-
-  getApiKey() {
-    const key = getEnvironmentVariable(this.envVar);
-    if (!key && !this.override) {
-      throw new Error(`Missing ${this.envVar} environment variable.`);
-    }
-    return key;
-  }
-
-  async geocodeCity(city) {
-    const geocodeUrl = `https://api.openweathermap.org/geo/1.0/direct?q=${encodeURIComponent(
-      city,
-    )}&limit=1&appid=${this.apiKey}`;
-    const res = await fetch(geocodeUrl);
-    const data = await res.json();
-    if (!res.ok || !Array.isArray(data) || data.length === 0) {
-      throw new Error(`Could not find coordinates for city: ${city}`);
-    }
-    return { lat: data[0].lat, lon: data[0].lon };
-  }
-
-  convertDateToUnix(dateStr) {
-    const parts = dateStr.split('-');
-    if (parts.length !== 3) {
-      throw new Error('Invalid date format. Expected YYYY-MM-DD.');
-    }
-    const year = parseInt(parts[0], 10);
-    const month = parseInt(parts[1], 10);
-    const day = parseInt(parts[2], 10);
-    if (isNaN(year) || isNaN(month) || isNaN(day)) {
-      throw new Error('Invalid date format. Expected YYYY-MM-DD with valid numbers.');
-    }
-
-    const dateObj = new Date(Date.UTC(year, month - 1, day, 0, 0, 0));
-    if (isNaN(dateObj.getTime())) {
-      throw new Error('Invalid date provided. Cannot parse into a valid date.');
-    }
-
-    return Math.floor(dateObj.getTime() / 1000);
-  }
-
-  async _call(args) {
-    try {
-      const { action, city, lat, lon, exclude, units, lang, date, tz } = args;
-      const owmUnits = mapUnitsToOpenWeather(units);
-
-      if (action === 'help') {
-        return JSON.stringify(
-          {
-            title: 'OpenWeather One Call API 3.0 Help',
-            description: 'Guidance on using the OpenWeather One Call API 3.0.',
-            endpoints: {
-              current_and_forecast: {
-                endpoint: 'data/3.0/onecall',
-                data_provided: [
-                  'Current weather',
-                  'Minute forecast (1h)',
-                  'Hourly forecast (48h)',
-                  'Daily forecast (8 days)',
-                  'Government weather alerts',
-                ],
-                required_params: [['lat', 'lon'], ['city']],
-                optional_params: ['exclude', 'units (Celsius/Kelvin/Fahrenheit)', 'lang'],
-                usage_example: {
-                  city: 'Knoxville, Tennessee',
-                  units: 'Fahrenheit',
-                  lang: 'en',
-                },
-              },
-              weather_for_timestamp: {
-                endpoint: 'data/3.0/onecall/timemachine',
-                data_provided: [
-                  'Historical weather (since 1979-01-01)',
-                  'Future forecast up to 4 days ahead',
-                ],
-                required_params: [
-                  ['lat', 'lon', 'date (YYYY-MM-DD)'],
-                  ['city', 'date (YYYY-MM-DD)'],
-                ],
-                optional_params: ['units (Celsius/Kelvin/Fahrenheit)', 'lang'],
-                usage_example: {
-                  city: 'Knoxville, Tennessee',
-                  date: '2020-03-04',
-                  units: 'Fahrenheit',
-                  lang: 'en',
-                },
-              },
-              daily_aggregation: {
-                endpoint: 'data/3.0/onecall/day_summary',
-                data_provided: [
-                  'Aggregated weather data for a specific date (1979-01-02 to 1.5 years ahead)',
-                ],
-                required_params: [
-                  ['lat', 'lon', 'date (YYYY-MM-DD)'],
-                  ['city', 'date (YYYY-MM-DD)'],
-                ],
-                optional_params: ['units (Celsius/Kelvin/Fahrenheit)', 'lang', 'tz'],
-                usage_example: {
-                  city: 'Knoxville, Tennessee',
-                  date: '2020-03-04',
-                  units: 'Celsius',
-                  lang: 'en',
-                },
-              },
-              weather_overview: {
-                endpoint: 'data/3.0/onecall/overview',
-                data_provided: ['Human-readable weather summary (today/tomorrow)'],
-                required_params: [['lat', 'lon'], ['city']],
-                optional_params: ['date (YYYY-MM-DD)', 'units (Celsius/Kelvin/Fahrenheit)'],
-                usage_example: {
-                  city: 'Knoxville, Tennessee',
-                  date: '2024-05-13',
-                  units: 'Celsius',
-                },
-              },
-            },
-            notes: [
-              'If lat/lon not provided, you can specify a city name and it will be geocoded.',
-              'For the timestamp action, provide a date in YYYY-MM-DD format instead of a Unix timestamp.',
-              'By default, temperatures are returned in Celsius.',
-              'You can specify units as Celsius, Kelvin, or Fahrenheit.',
-              'All temperatures are rounded to the nearest degree.',
-            ],
-            errors: [
-              '400: Bad Request (missing/invalid params)',
-              '401: Unauthorized (check API key)',
-              '404: Not Found (no data or city)',
-              '429: Too many requests',
-              '5xx: Internal error',
-            ],
-          },
-          null,
-          2,
-        );
-      }
-
-      let finalLat = lat;
-      let finalLon = lon;
-
-      // If lat/lon not provided but city is given, geocode it
-      if ((finalLat == null || finalLon == null) && city) {
-        const coords = await this.geocodeCity(city);
-        finalLat = coords.lat;
-        finalLon = coords.lon;
-      }
-
-      if (['current_forecast', 'timestamp', 'daily_aggregation', 'overview'].includes(action)) {
-        if (typeof finalLat !== 'number' || typeof finalLon !== 'number') {
-          return 'Error: lat and lon are required and must be numbers for this action (or specify \'city\').';
-        }
-      }
-
-      const baseUrl = 'https://api.openweathermap.org/data/3.0';
-      let endpoint = '';
-      const params = new URLSearchParams({ appid: this.apiKey, units: owmUnits });
-
-      let dt;
-      if (action === 'timestamp') {
-        if (!date) {
-          return 'Error: For timestamp action, a \'date\' in YYYY-MM-DD format is required.';
-        }
-        dt = this.convertDateToUnix(date);
-      }
-
-      if (action === 'daily_aggregation' && !date) {
-        return 'Error: date (YYYY-MM-DD) is required for daily_aggregation action.';
-      }
-
-      switch (action) {
-        case 'current_forecast':
-          endpoint = '/onecall';
-          params.append('lat', String(finalLat));
-          params.append('lon', String(finalLon));
-          if (exclude) {
-            params.append('exclude', exclude);
-          }
-          if (lang) {
-            params.append('lang', lang);
-          }
-          break;
-        case 'timestamp':
-          endpoint = '/onecall/timemachine';
-          params.append('lat', String(finalLat));
-          params.append('lon', String(finalLon));
-          params.append('dt', String(dt));
-          if (lang) {
-            params.append('lang', lang);
-          }
-          break;
-        case 'daily_aggregation':
-          endpoint = '/onecall/day_summary';
-          params.append('lat', String(finalLat));
-          params.append('lon', String(finalLon));
-          params.append('date', date);
-          if (lang) {
-            params.append('lang', lang);
-          }
-          if (tz) {
-            params.append('tz', tz);
-          }
-          break;
-        case 'overview':
-          endpoint = '/onecall/overview';
-          params.append('lat', String(finalLat));
-          params.append('lon', String(finalLon));
-          if (date) {
-            params.append('date', date);
-          }
-          break;
-        default:
-          return `Error: Unknown action: ${action}`;
-      }
-
-      const url = `${baseUrl}${endpoint}?${params.toString()}`;
-      const response = await fetch(url);
-      const json = await response.json();
-      if (!response.ok) {
-        return `Error: OpenWeather API request failed with status ${response.status}: ${
-          json.message || JSON.stringify(json)
-        }`;
-      }
-
-      const roundedJson = roundTemperatures(json);
-      return JSON.stringify(roundedJson);
-    } catch (err) {
-      return `Error: ${err.message}`;
-    }
-  }
-}
-
-module.exports = OpenWeather;
--- a/api/app/clients/tools/structured/StableDiffusion.js
+++ b/api/app/clients/tools/structured/StableDiffusion.js
@@ -6,26 +6,21 @@ const axios = require('axios');
 const sharp = require('sharp');
 const { v4: uuidv4 } = require('uuid');
 const { Tool } = require('@langchain/core/tools');
-const { FileContext, ContentTypes } = require('librechat-data-provider');
+const { FileContext } = require('librechat-data-provider');
 const paths = require('~/config/paths');
 const { logger } = require('~/config');

-const displayMessage =
-  "Stable Diffusion displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
-
 class StableDiffusionAPI extends Tool {
  constructor(fields) {
    super();
    /** @type {string} User ID */
    this.userId = fields.userId;
-    /** @type {ServerRequest | undefined} Express Request object, only provided by ToolService */
+    /** @type {Express.Request | undefined} Express Request object, only provided by ToolService */
    this.req = fields.req;
    /** @type {boolean} Used to initialize the Tool without necessary variables. */
    this.override = fields.override ?? false;
    /** @type {boolean} Necessary for output to contain all image metadata. */
    this.returnMetadata = fields.returnMetadata ?? false;
-    /** @type {boolean} */
-    this.isAgent = fields.isAgent;
    if (fields.uploadImageBuffer) {
      /** @type {uploadImageBuffer} Necessary for output to contain all image metadata. */
      this.uploadImageBuffer = fields.uploadImageBuffer.bind(this);
@@ -44,7 +39,7 @@ class StableDiffusionAPI extends Tool {
 // "negative_prompt":"semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, out of frame, low quality, ugly, mutation, deformed"
 // - Generate images only once per human query unless explicitly requested by the user`;
    this.description =
-      "You can generate images using text with 'stable-diffusion'. This tool is exclusively for visual content.";
+      'You can generate images using text with \'stable-diffusion\'. This tool is exclusively for visual content.';
    this.schema = z.object({
      prompt: z
        .string()
@@ -71,16 +66,6 @@ class StableDiffusionAPI extends Tool {
    return `![generated image](/${imageUrl})`;
  }

-  returnValue(value) {
-    if (this.isAgent === true && typeof value === 'string') {
-      return [value, {}];
-    } else if (this.isAgent === true && typeof value === 'object') {
-      return [displayMessage, value];
-    }
-
-    return value;
-  }
-
  getServerURL() {
    const url = process.env.SD_WEBUI_URL || '';
    if (!url && !this.override) {
@@ -128,25 +113,6 @@ class StableDiffusionAPI extends Tool {
    }

    try {
-      if (this.isAgent) {
-        const content = [
-          {
-            type: ContentTypes.IMAGE_URL,
-            image_url: {
-              url: `data:image/png;base64,${image}`,
-            },
-          },
-        ];
-
-        const response = [
-          {
-            type: ContentTypes.TEXT,
-            text: displayMessage,
-          },
-        ];
-        return [response, { content }];
-      }
-
      const buffer = Buffer.from(image.split(',', 1)[0], 'base64');
      if (this.returnMetadata && this.uploadImageBuffer && this.req) {
        const file = await this.uploadImageBuffer({
@@ -188,7 +154,7 @@ class StableDiffusionAPI extends Tool {
      logger.error('[StableDiffusion] Error while saving the image:', error);
    }

-    return this.returnValue(this.result);
+    return this.result;
  }
 }

--- a/api/app/clients/tools/structured/TavilySearch.js
+++ b/api/app/clients/tools/structured/TavilySearch.js
@@ -1,6 +1,6 @@
 const { z } = require('zod');
 const { tool } = require('@langchain/core/tools');
-const { getApiKey } = require('./credentials');
+const { getEnvironmentVariable } = require('@langchain/core/utils/env');

 function createTavilySearchTool(fields = {}) {
  const envVar = 'TAVILY_API_KEY';
@@ -8,6 +8,14 @@ function createTavilySearchTool(fields = {}) {
  const apiKey = fields.apiKey ?? getApiKey(envVar, override);
  const kwargs = fields?.kwargs ?? {};

+  function getApiKey(envVar, override) {
+    const key = getEnvironmentVariable(envVar);
+    if (!key && !override) {
+      throw new Error(`Missing ${envVar} environment variable.`);
+    }
+    return key;
+  }
+
  return tool(
    async (input) => {
      const { query, ...rest } = input;
--- a/api/app/clients/tools/structured/TavilySearchResults.js
+++ b/api/app/clients/tools/structured/TavilySearchResults.js
@@ -43,39 +43,9 @@ class TavilySearchResults extends Tool {
        .boolean()
        .optional()
        .describe('Whether to include answers in the search results. Default is False.'),
-      include_raw_content: z
-        .boolean()
-        .optional()
-        .describe('Whether to include raw content in the search results. Default is False.'),
-      include_domains: z
-        .array(z.string())
-        .optional()
-        .describe('A list of domains to specifically include in the search results.'),
-      exclude_domains: z
-        .array(z.string())
-        .optional()
-        .describe('A list of domains to specifically exclude from the search results.'),
-      topic: z
-        .enum(['general', 'news', 'finance'])
-        .optional()
-        .describe(
-          'The category of the search. Use news ONLY if query SPECIFCALLY mentions the word "news".',
-        ),
-      time_range: z
-        .enum(['day', 'week', 'month', 'year', 'd', 'w', 'm', 'y'])
-        .optional()
-        .describe('The time range back from the current date to filter results.'),
-      days: z
-        .number()
-        .min(1)
-        .optional()
-        .describe('Number of days back from the current date to include. Only if topic is news.'),
-      include_image_descriptions: z
-        .boolean()
-        .optional()
-        .describe(
-          'When include_images is true, also add a descriptive text for each image. Default is false.',
-        ),
+      // include_raw_content: z.boolean().optional().describe('Whether to include raw content in the search results. Default is False.'),
+      // include_domains: z.array(z.string()).optional().describe('A list of domains to specifically include in the search results.'),
+      // exclude_domains: z.array(z.string()).optional().describe('A list of domains to specifically exclude from the search results.'),
    });
  }

--- a/api/app/clients/tools/structured/YouTube.js
+++ b/api/app/clients/tools/structured/YouTube.js
@@ -1,137 +0,0 @@
-const { ytToolkit } = require('@librechat/api');
-const { tool } = require('@langchain/core/tools');
-const { youtube } = require('@googleapis/youtube');
-const { logger } = require('@librechat/data-schemas');
-const { YoutubeTranscript } = require('youtube-transcript');
-const { getApiKey } = require('./credentials');
-
-function extractVideoId(url) {
-  const rawIdRegex = /^[a-zA-Z0-9_-]{11}$/;
-  if (rawIdRegex.test(url)) {
-    return url;
-  }
-
-  const regex = new RegExp(
-    '(?:youtu\\.be/|youtube(?:\\.com)?/(?:' +
-      '(?:watch\\?v=)|(?:embed/)|(?:shorts/)|(?:live/)|(?:v/)|(?:/))?)' +
-      '([a-zA-Z0-9_-]{11})(?:\\S+)?$',
-  );
-  const match = url.match(regex);
-  return match ? match[1] : null;
-}
-
-function parseTranscript(transcriptResponse) {
-  if (!Array.isArray(transcriptResponse)) {
-    return '';
-  }
-
-  return transcriptResponse
-    .map((entry) => entry.text.trim())
-    .filter((text) => text)
-    .join(' ')
-    .replaceAll('&amp;#39;', "'");
-}
-
-function createYouTubeTools(fields = {}) {
-  const envVar = 'YOUTUBE_API_KEY';
-  const override = fields.override ?? false;
-  const apiKey = fields.apiKey ?? fields[envVar] ?? getApiKey(envVar, override);
-
-  const youtubeClient = youtube({
-    version: 'v3',
-    auth: apiKey,
-  });
-
-  const searchTool = tool(async ({ query, maxResults = 5 }) => {
-    const response = await youtubeClient.search.list({
-      part: 'snippet',
-      q: query,
-      type: 'video',
-      maxResults: maxResults || 5,
-    });
-    const result = response.data.items.map((item) => ({
-      title: item.snippet.title,
-      description: item.snippet.description,
-      url: `https://www.youtube.com/watch?v=${item.id.videoId}`,
-    }));
-    return JSON.stringify(result, null, 2);
-  }, ytToolkit.youtube_search);
-
-  const infoTool = tool(async ({ url }) => {
-    const videoId = extractVideoId(url);
-    if (!videoId) {
-      throw new Error('Invalid YouTube URL or video ID');
-    }
-
-    const response = await youtubeClient.videos.list({
-      part: 'snippet,statistics',
-      id: videoId,
-    });
-
-    if (!response.data.items?.length) {
-      throw new Error('Video not found');
-    }
-    const video = response.data.items[0];
-
-    const result = {
-      title: video.snippet.title,
-      description: video.snippet.description,
-      views: video.statistics.viewCount,
-      likes: video.statistics.likeCount,
-      comments: video.statistics.commentCount,
-    };
-    return JSON.stringify(result, null, 2);
-  }, ytToolkit.youtube_info);
-
-  const commentsTool = tool(async ({ url, maxResults = 10 }) => {
-    const videoId = extractVideoId(url);
-    if (!videoId) {
-      throw new Error('Invalid YouTube URL or video ID');
-    }
-
-    const response = await youtubeClient.commentThreads.list({
-      part: 'snippet',
-      videoId,
-      maxResults: maxResults || 10,
-    });
-
-    const result = response.data.items.map((item) => ({
-      author: item.snippet.topLevelComment.snippet.authorDisplayName,
-      text: item.snippet.topLevelComment.snippet.textDisplay,
-      likes: item.snippet.topLevelComment.snippet.likeCount,
-    }));
-    return JSON.stringify(result, null, 2);
-  }, ytToolkit.youtube_comments);
-
-  const transcriptTool = tool(async ({ url }) => {
-    const videoId = extractVideoId(url);
-    if (!videoId) {
-      throw new Error('Invalid YouTube URL or video ID');
-    }
-
-    try {
-      try {
-        const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'en' });
-        return parseTranscript(transcript);
-      } catch (e) {
-        logger.error(e);
-      }
-
-      try {
-        const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'de' });
-        return parseTranscript(transcript);
-      } catch (e) {
-        logger.error(e);
-      }
-
-      const transcript = await YoutubeTranscript.fetchTranscript(videoId);
-      return parseTranscript(transcript);
-    } catch (error) {
-      throw new Error(`Failed to fetch transcript: ${error.message}`);
-    }
-  }, ytToolkit.youtube_transcript);
-
-  return [searchTool, infoTool, commentsTool, transcriptTool];
-}
-
-module.exports = createYouTubeTools;
--- a/api/app/clients/tools/structured/credentials.js
+++ b/api/app/clients/tools/structured/credentials.js
@@ -1,13 +0,0 @@
-const { getEnvironmentVariable } = require('@langchain/core/utils/env');
-
-function getApiKey(envVar, override) {
-  const key = getEnvironmentVariable(envVar);
-  if (!key && !override) {
-    throw new Error(`Missing ${envVar} environment variable.`);
-  }
-  return key;
-}
-
-module.exports = {
-  getApiKey,
-};
--- a/api/app/clients/tools/structured/specs/DALLE3-proxy.spec.js
+++ b/api/app/clients/tools/structured/specs/DALLE3-proxy.spec.js
@@ -1,60 +0,0 @@
-const DALLE3 = require('../DALLE3');
-const { ProxyAgent } = require('undici');
-
-jest.mock('tiktoken');
-const processFileURL = jest.fn();
-
-describe('DALLE3 Proxy Configuration', () => {
-  let originalEnv;
-
-  beforeAll(() => {
-    originalEnv = { ...process.env };
-  });
-
-  beforeEach(() => {
-    jest.resetModules();
-    process.env = { ...originalEnv };
-  });
-
-  afterEach(() => {
-    process.env = originalEnv;
-  });
-
-  it('should configure ProxyAgent in fetchOptions.dispatcher when PROXY env is set', () => {
-    // Set proxy environment variable
-    process.env.PROXY = 'http://proxy.example.com:8080';
-    process.env.DALLE_API_KEY = 'test-api-key';
-
-    // Create instance
-    const dalleWithProxy = new DALLE3({ processFileURL });
-
-    // Check that the openai client exists
-    expect(dalleWithProxy.openai).toBeDefined();
-
-    // Check that _options exists and has fetchOptions with a dispatcher
-    expect(dalleWithProxy.openai._options).toBeDefined();
-    expect(dalleWithProxy.openai._options.fetchOptions).toBeDefined();
-    expect(dalleWithProxy.openai._options.fetchOptions.dispatcher).toBeDefined();
-    expect(dalleWithProxy.openai._options.fetchOptions.dispatcher).toBeInstanceOf(ProxyAgent);
-  });
-
-  it('should not configure ProxyAgent when PROXY env is not set', () => {
-    // Ensure PROXY is not set
-    delete process.env.PROXY;
-    process.env.DALLE_API_KEY = 'test-api-key';
-
-    // Create instance
-    const dalleWithoutProxy = new DALLE3({ processFileURL });
-
-    // Check that the openai client exists
-    expect(dalleWithoutProxy.openai).toBeDefined();
-
-    // Check that _options exists but fetchOptions either doesn't exist or doesn't have a dispatcher
-    expect(dalleWithoutProxy.openai._options).toBeDefined();
-
-    // fetchOptions should either not exist or not have a dispatcher
-    if (dalleWithoutProxy.openai._options.fetchOptions) {
-      expect(dalleWithoutProxy.openai._options.fetchOptions.dispatcher).toBeUndefined();
-    }
-  });
-});
--- a/api/app/clients/tools/structured/specs/DALLE3.spec.js
+++ b/api/app/clients/tools/structured/specs/DALLE3.spec.js
@@ -1,30 +1,31 @@
 const OpenAI = require('openai');
-const { logger } = require('@librechat/data-schemas');
 const DALLE3 = require('../DALLE3');

-jest.mock('openai');
-jest.mock('@librechat/data-schemas', () => {
-  return {
-    logger: {
-      info: jest.fn(),
-      warn: jest.fn(),
-      debug: jest.fn(),
-      error: jest.fn(),
-    },
-  };
-});
+const { logger } = require('~/config');

-jest.mock('tiktoken', () => {
-  return {
-    encoding_for_model: jest.fn().mockReturnValue({
-      encode: jest.fn(),
-      decode: jest.fn(),
-    }),
-  };
-});
+jest.mock('openai');

 const processFileURL = jest.fn();

+jest.mock('~/server/services/Files/images', () => ({
+  getImageBasename: jest.fn().mockImplementation((url) => {
+    // Split the URL by '/'
+    const parts = url.split('/');
+
+    // Get the last part of the URL
+    const lastPart = parts.pop();
+
+    // Check if the last part of the URL matches the image extension regex
+    const imageExtensionRegex = /\.(jpg|jpeg|png|gif|bmp|tiff|svg)$/i;
+    if (imageExtensionRegex.test(lastPart)) {
+      return lastPart;
+    }
+
+    // If the regex test fails, return an empty string
+    return '';
+  }),
+}));
+
 const generate = jest.fn();
 OpenAI.mockImplementation(() => ({
  images: {
@@ -36,11 +37,6 @@ jest.mock('fs', () => {
  return {
    existsSync: jest.fn(),
    mkdirSync: jest.fn(),
-    promises: {
-      writeFile: jest.fn(),
-      readFile: jest.fn(),
-      unlink: jest.fn(),
-    },
  };
 });

--- a/api/app/clients/tools/structured/specs/openWeather.integration.test.js
+++ b/api/app/clients/tools/structured/specs/openWeather.integration.test.js
@@ -1,224 +0,0 @@
-// __tests__/openWeather.integration.test.js
-const OpenWeather = require('../OpenWeather');
-
-describe('OpenWeather Tool (Integration Test)', () => {
-  let tool;
-
-  beforeAll(() => {
-    tool = new OpenWeather({ override: true });
-    console.log('API Key present:', !!process.env.OPENWEATHER_API_KEY);
-  });
-
-  test('current_forecast with a real API key returns current weather', async () => {
-    // Check if API key is available
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      const result = await tool.call({
-        action: 'current_forecast',
-        city: 'London',
-        units: 'Celsius',
-      });
-
-      console.log('Raw API response:', result);
-
-      const parsed = JSON.parse(result);
-      expect(parsed).toHaveProperty('current');
-      expect(typeof parsed.current.temp).toBe('number');
-    } catch (error) {
-      console.error('Test failed with error:', error);
-      throw error;
-    }
-  });
-
-  test('timestamp action with real API key returns historical data', async () => {
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      // Use a date from yesterday to ensure data availability
-      const yesterday = new Date();
-      yesterday.setDate(yesterday.getDate() - 1);
-      const dateStr = yesterday.toISOString().split('T')[0];
-
-      const result = await tool.call({
-        action: 'timestamp',
-        city: 'London',
-        date: dateStr,
-        units: 'Celsius',
-      });
-
-      console.log('Timestamp API response:', result);
-
-      const parsed = JSON.parse(result);
-      expect(parsed).toHaveProperty('data');
-      expect(Array.isArray(parsed.data)).toBe(true);
-      expect(parsed.data[0]).toHaveProperty('temp');
-    } catch (error) {
-      console.error('Timestamp test failed with error:', error);
-      throw error;
-    }
-  });
-
-  test('daily_aggregation action with real API key returns aggregated data', async () => {
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      // Use yesterday's date for aggregation
-      const yesterday = new Date();
-      yesterday.setDate(yesterday.getDate() - 1);
-      const dateStr = yesterday.toISOString().split('T')[0];
-
-      const result = await tool.call({
-        action: 'daily_aggregation',
-        city: 'London',
-        date: dateStr,
-        units: 'Celsius',
-      });
-
-      console.log('Daily aggregation API response:', result);
-
-      const parsed = JSON.parse(result);
-      expect(parsed).toHaveProperty('temperature');
-      expect(parsed.temperature).toHaveProperty('morning');
-      expect(parsed.temperature).toHaveProperty('afternoon');
-      expect(parsed.temperature).toHaveProperty('evening');
-    } catch (error) {
-      console.error('Daily aggregation test failed with error:', error);
-      throw error;
-    }
-  });
-
-  test('overview action with real API key returns weather summary', async () => {
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      const result = await tool.call({
-        action: 'overview',
-        city: 'London',
-        units: 'Celsius',
-      });
-
-      console.log('Overview API response:', result);
-
-      const parsed = JSON.parse(result);
-      expect(parsed).toHaveProperty('weather_overview');
-      expect(typeof parsed.weather_overview).toBe('string');
-      expect(parsed.weather_overview.length).toBeGreaterThan(0);
-      expect(parsed).toHaveProperty('date');
-      expect(parsed).toHaveProperty('units');
-      expect(parsed.units).toBe('metric');
-    } catch (error) {
-      console.error('Overview test failed with error:', error);
-      throw error;
-    }
-  });
-
-  test('different temperature units return correct values', async () => {
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      // Test Celsius
-      let result = await tool.call({
-        action: 'current_forecast',
-        city: 'London',
-        units: 'Celsius',
-      });
-      let parsed = JSON.parse(result);
-      const celsiusTemp = parsed.current.temp;
-
-      // Test Kelvin
-      result = await tool.call({
-        action: 'current_forecast',
-        city: 'London',
-        units: 'Kelvin',
-      });
-      parsed = JSON.parse(result);
-      const kelvinTemp = parsed.current.temp;
-
-      // Test Fahrenheit
-      result = await tool.call({
-        action: 'current_forecast',
-        city: 'London',
-        units: 'Fahrenheit',
-      });
-      parsed = JSON.parse(result);
-      const fahrenheitTemp = parsed.current.temp;
-
-      // Verify temperature conversions are roughly correct
-      // K = C + 273.15
-      // F = (C * 9/5) + 32
-      const celsiusToKelvin = Math.round(celsiusTemp + 273.15);
-      const celsiusToFahrenheit = Math.round((celsiusTemp * 9) / 5 + 32);
-
-      console.log('Temperature comparisons:', {
-        celsius: celsiusTemp,
-        kelvin: kelvinTemp,
-        fahrenheit: fahrenheitTemp,
-        calculatedKelvin: celsiusToKelvin,
-        calculatedFahrenheit: celsiusToFahrenheit,
-      });
-
-      // Allow for some rounding differences
-      expect(Math.abs(kelvinTemp - celsiusToKelvin)).toBeLessThanOrEqual(1);
-      expect(Math.abs(fahrenheitTemp - celsiusToFahrenheit)).toBeLessThanOrEqual(1);
-    } catch (error) {
-      console.error('Temperature units test failed with error:', error);
-      throw error;
-    }
-  });
-
-  test('language parameter returns localized data', async () => {
-    if (!process.env.OPENWEATHER_API_KEY) {
-      console.warn('Skipping real API test, no OPENWEATHER_API_KEY found.');
-      return;
-    }
-
-    try {
-      // Test with English
-      let result = await tool.call({
-        action: 'current_forecast',
-        city: 'Paris',
-        units: 'Celsius',
-        lang: 'en',
-      });
-      let parsed = JSON.parse(result);
-      const englishDescription = parsed.current.weather[0].description;
-
-      // Test with French
-      result = await tool.call({
-        action: 'current_forecast',
-        city: 'Paris',
-        units: 'Celsius',
-        lang: 'fr',
-      });
-      parsed = JSON.parse(result);
-      const frenchDescription = parsed.current.weather[0].description;
-
-      console.log('Language comparison:', {
-        english: englishDescription,
-        french: frenchDescription,
-      });
-
-      // Verify descriptions are different (indicating translation worked)
-      expect(englishDescription).not.toBe(frenchDescription);
-    } catch (error) {
-      console.error('Language test failed with error:', error);
-      throw error;
-    }
-  });
-});
--- a/api/app/clients/tools/structured/specs/openweather.test.js
+++ b/api/app/clients/tools/structured/specs/openweather.test.js
@@ -1,358 +0,0 @@
-// __tests__/openweather.test.js
-const OpenWeather = require('../OpenWeather');
-const fetch = require('node-fetch');
-
-// Mock environment variable
-process.env.OPENWEATHER_API_KEY = 'test-api-key';
-
-// Mock the fetch function globally
-jest.mock('node-fetch', () => jest.fn());
-
-describe('OpenWeather Tool', () => {
-  let tool;
-
-  beforeAll(() => {
-    tool = new OpenWeather();
-  });
-
-  beforeEach(() => {
-    fetch.mockReset();
-  });
-
-  test('action=help returns help instructions', async () => {
-    const result = await tool.call({
-      action: 'help',
-    });
-
-    expect(typeof result).toBe('string');
-    const parsed = JSON.parse(result);
-    expect(parsed.title).toBe('OpenWeather One Call API 3.0 Help');
-  });
-
-  test('current_forecast with a city and successful geocoding + forecast', async () => {
-    // Mock geocoding response
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('geo/1.0/direct')) {
-        return Promise.resolve({
-          ok: true,
-          json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-        });
-      }
-      return Promise.reject('Unexpected fetch call for geocoding');
-    });
-
-    // Mock forecast response
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => ({
-          current: { temp: 293.15, feels_like: 295.15 },
-          daily: [{ temp: { day: 293.15, night: 283.15 } }],
-        }),
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'current_forecast',
-      city: 'Knoxville, Tennessee',
-      units: 'Kelvin',
-    });
-
-    const parsed = JSON.parse(result);
-    expect(parsed.current.temp).toBe(293);
-    expect(parsed.current.feels_like).toBe(295);
-    expect(parsed.daily[0].temp.day).toBe(293);
-    expect(parsed.daily[0].temp.night).toBe(283);
-  });
-
-  test('timestamp action with valid date returns mocked historical data', async () => {
-    // Mock geocoding response
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('geo/1.0/direct')) {
-        return Promise.resolve({
-          ok: true,
-          json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-        });
-      }
-      return Promise.reject('Unexpected fetch call for geocoding');
-    });
-
-    // Mock historical weather response
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => ({
-          data: [
-            {
-              dt: 1583280000,
-              temp: 283.15,
-              feels_like: 280.15,
-              humidity: 75,
-              weather: [{ description: 'clear sky' }],
-            },
-          ],
-        }),
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'timestamp',
-      city: 'Knoxville, Tennessee',
-      date: '2020-03-04',
-      units: 'Kelvin',
-    });
-
-    const parsed = JSON.parse(result);
-    expect(parsed.data[0].temp).toBe(283);
-    expect(parsed.data[0].feels_like).toBe(280);
-  });
-
-  test('daily_aggregation action returns aggregated weather data', async () => {
-    // Mock geocoding response
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('geo/1.0/direct')) {
-        return Promise.resolve({
-          ok: true,
-          json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-        });
-      }
-      return Promise.reject('Unexpected fetch call for geocoding');
-    });
-
-    // Mock daily aggregation response
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => ({
-          date: '2020-03-04',
-          temperature: {
-            morning: 283.15,
-            afternoon: 293.15,
-            evening: 288.15,
-          },
-          humidity: {
-            morning: 75,
-            afternoon: 60,
-            evening: 70,
-          },
-        }),
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'daily_aggregation',
-      city: 'Knoxville, Tennessee',
-      date: '2020-03-04',
-      units: 'Kelvin',
-    });
-
-    const parsed = JSON.parse(result);
-    expect(parsed.temperature.morning).toBe(283);
-    expect(parsed.temperature.afternoon).toBe(293);
-    expect(parsed.temperature.evening).toBe(288);
-  });
-
-  test('overview action returns weather summary', async () => {
-    // Mock geocoding response
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('geo/1.0/direct')) {
-        return Promise.resolve({
-          ok: true,
-          json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-        });
-      }
-      return Promise.reject('Unexpected fetch call for geocoding');
-    });
-
-    // Mock overview response
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => ({
-          date: '2024-01-07',
-          lat: 35.9606,
-          lon: -83.9207,
-          tz: '+00:00',
-          units: 'metric',
-          weather_overview:
-            'Currently, the temperature is 2°C with a real feel of -2°C. The sky is clear with moderate wind.',
-        }),
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'overview',
-      city: 'Knoxville, Tennessee',
-      units: 'Celsius',
-    });
-
-    const parsed = JSON.parse(result);
-    expect(parsed).toHaveProperty('weather_overview');
-    expect(typeof parsed.weather_overview).toBe('string');
-    expect(parsed.weather_overview.length).toBeGreaterThan(0);
-    expect(parsed).toHaveProperty('date');
-    expect(parsed).toHaveProperty('units');
-    expect(parsed.units).toBe('metric');
-  });
-
-  test('temperature units are correctly converted', async () => {
-    // Mock geocoding response for all three calls
-    const geocodingMock = Promise.resolve({
-      ok: true,
-      json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-    });
-
-    // Mock weather response for Kelvin
-    const kelvinMock = Promise.resolve({
-      ok: true,
-      json: async () => ({
-        current: { temp: 293.15 },
-      }),
-    });
-
-    // Mock weather response for Celsius
-    const celsiusMock = Promise.resolve({
-      ok: true,
-      json: async () => ({
-        current: { temp: 20 },
-      }),
-    });
-
-    // Mock weather response for Fahrenheit
-    const fahrenheitMock = Promise.resolve({
-      ok: true,
-      json: async () => ({
-        current: { temp: 68 },
-      }),
-    });
-
-    // Test Kelvin
-    fetch.mockImplementationOnce(() => geocodingMock).mockImplementationOnce(() => kelvinMock);
-
-    let result = await tool.call({
-      action: 'current_forecast',
-      city: 'Knoxville, Tennessee',
-      units: 'Kelvin',
-    });
-    let parsed = JSON.parse(result);
-    expect(parsed.current.temp).toBe(293);
-
-    // Test Celsius
-    fetch.mockImplementationOnce(() => geocodingMock).mockImplementationOnce(() => celsiusMock);
-
-    result = await tool.call({
-      action: 'current_forecast',
-      city: 'Knoxville, Tennessee',
-      units: 'Celsius',
-    });
-    parsed = JSON.parse(result);
-    expect(parsed.current.temp).toBe(20);
-
-    // Test Fahrenheit
-    fetch.mockImplementationOnce(() => geocodingMock).mockImplementationOnce(() => fahrenheitMock);
-
-    result = await tool.call({
-      action: 'current_forecast',
-      city: 'Knoxville, Tennessee',
-      units: 'Fahrenheit',
-    });
-    parsed = JSON.parse(result);
-    expect(parsed.current.temp).toBe(68);
-  });
-
-  test('timestamp action without a date returns an error message', async () => {
-    const result = await tool.call({
-      action: 'timestamp',
-      lat: 35.9606,
-      lon: -83.9207,
-    });
-    expect(result).toMatch(
-      /Error: For timestamp action, a 'date' in YYYY-MM-DD format is required./,
-    );
-  });
-
-  test('daily_aggregation action without a date returns an error message', async () => {
-    const result = await tool.call({
-      action: 'daily_aggregation',
-      lat: 35.9606,
-      lon: -83.9207,
-    });
-    expect(result).toMatch(/Error: date \(YYYY-MM-DD\) is required for daily_aggregation action./);
-  });
-
-  test('unknown action returns an error due to schema validation', async () => {
-    await expect(
-      tool.call({
-        action: 'unknown_action',
-      }),
-    ).rejects.toThrow(/Received tool input did not match expected schema/);
-  });
-
-  test('geocoding failure returns a descriptive error', async () => {
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => [],
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'current_forecast',
-      city: 'NowhereCity',
-    });
-    expect(result).toMatch(/Error: Could not find coordinates for city: NowhereCity/);
-  });
-
-  test('API request failure returns an error', async () => {
-    // Mock geocoding success
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: true,
-        json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-      }),
-    );
-
-    // Mock weather request failure
-    fetch.mockImplementationOnce(() =>
-      Promise.resolve({
-        ok: false,
-        status: 404,
-        json: async () => ({ message: 'Not found' }),
-      }),
-    );
-
-    const result = await tool.call({
-      action: 'current_forecast',
-      city: 'Knoxville, Tennessee',
-    });
-    expect(result).toMatch(/Error: OpenWeather API request failed with status 404: Not found/);
-  });
-
-  test('invalid date format returns an error', async () => {
-    // Mock geocoding response first
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('geo/1.0/direct')) {
-        return Promise.resolve({
-          ok: true,
-          json: async () => [{ lat: 35.9606, lon: -83.9207 }],
-        });
-      }
-      return Promise.reject('Unexpected fetch call for geocoding');
-    });
-
-    // Mock timestamp API response
-    fetch.mockImplementationOnce((url) => {
-      if (url.includes('onecall/timemachine')) {
-        throw new Error('Invalid date format. Expected YYYY-MM-DD.');
-      }
-      return Promise.reject('Unexpected fetch call');
-    });
-
-    const result = await tool.call({
-      action: 'timestamp',
-      city: 'Knoxville, Tennessee',
-      date: '03-04-2020', // Wrong format
-    });
-    expect(result).toMatch(/Error: Invalid date format. Expected YYYY-MM-DD./);
-  });
-});
--- a/api/app/clients/tools/util/addOpenAPISpecs.js
+++ b/api/app/clients/tools/util/addOpenAPISpecs.js
@@ -0,0 +1,30 @@
+const { loadSpecs } = require('./loadSpecs');
+
+function transformSpec(input) {
+  return {
+    name: input.name_for_human,
+    pluginKey: input.name_for_model,
+    description: input.description_for_human,
+    icon: input?.logo_url ?? 'https://placehold.co/70x70.png',
+    // TODO: add support for authentication
+    isAuthRequired: 'false',
+    authConfig: [],
+  };
+}
+
+async function addOpenAPISpecs(availableTools) {
+  try {
+    const specs = (await loadSpecs({})).map(transformSpec);
+    if (specs.length > 0) {
+      return [...specs, ...availableTools];
+    }
+    return availableTools;
+  } catch (error) {
+    return availableTools;
+  }
+}
+
+module.exports = {
+  transformSpec,
+  addOpenAPISpecs,
+};
--- a/api/app/clients/tools/util/addOpenAPISpecs.spec.js
+++ b/api/app/clients/tools/util/addOpenAPISpecs.spec.js
@@ -0,0 +1,76 @@
+const { addOpenAPISpecs, transformSpec } = require('./addOpenAPISpecs');
+const { loadSpecs } = require('./loadSpecs');
+const { createOpenAPIPlugin } = require('../dynamic/OpenAPIPlugin');
+
+jest.mock('./loadSpecs');
+jest.mock('../dynamic/OpenAPIPlugin');
+
+describe('transformSpec', () => {
+  it('should transform input spec to a desired format', () => {
+    const input = {
+      name_for_human: 'Human Name',
+      name_for_model: 'Model Name',
+      description_for_human: 'Human Description',
+      logo_url: 'https://example.com/logo.png',
+    };
+
+    const expectedOutput = {
+      name: 'Human Name',
+      pluginKey: 'Model Name',
+      description: 'Human Description',
+      icon: 'https://example.com/logo.png',
+      isAuthRequired: 'false',
+      authConfig: [],
+    };
+
+    expect(transformSpec(input)).toEqual(expectedOutput);
+  });
+
+  it('should use default icon if logo_url is not provided', () => {
+    const input = {
+      name_for_human: 'Human Name',
+      name_for_model: 'Model Name',
+      description_for_human: 'Human Description',
+    };
+
+    const expectedOutput = {
+      name: 'Human Name',
+      pluginKey: 'Model Name',
+      description: 'Human Description',
+      icon: 'https://placehold.co/70x70.png',
+      isAuthRequired: 'false',
+      authConfig: [],
+    };
+
+    expect(transformSpec(input)).toEqual(expectedOutput);
+  });
+});
+
+describe('addOpenAPISpecs', () => {
+  it('should add specs to available tools', async () => {
+    const availableTools = ['Tool1', 'Tool2'];
+    const specs = [
+      {
+        name_for_human: 'Human Name',
+        name_for_model: 'Model Name',
+        description_for_human: 'Human Description',
+        logo_url: 'https://example.com/logo.png',
+      },
+    ];
+
+    loadSpecs.mockResolvedValue(specs);
+    createOpenAPIPlugin.mockReturnValue('Plugin');
+
+    const result = await addOpenAPISpecs(availableTools);
+    expect(result).toEqual([...specs.map(transformSpec), ...availableTools]);
+  });
+
+  it('should return available tools if specs loading fails', async () => {
+    const availableTools = ['Tool1', 'Tool2'];
+
+    loadSpecs.mockRejectedValue(new Error('Failed to load specs'));
+
+    const result = await addOpenAPISpecs(availableTools);
+    expect(result).toEqual(availableTools);
+  });
+});
--- a/api/app/clients/tools/util/createFileSearchTool.js
+++ b/api/app/clients/tools/util/createFileSearchTool.js
@@ -0,0 +1,104 @@
+const { z } = require('zod');
+const axios = require('axios');
+const { tool } = require('@langchain/core/tools');
+const { Tools, EToolResources } = require('librechat-data-provider');
+const { getFiles } = require('~/models/File');
+const { logger } = require('~/config');
+
+/**
+ *
+ * @param {Object} options
+ * @param {ServerRequest} options.req
+ * @param {Agent['tool_resources']} options.tool_resources
+ * @returns
+ */
+const createFileSearchTool = async (options) => {
+  const { req, tool_resources } = options;
+  const file_ids = tool_resources?.[EToolResources.file_search]?.file_ids ?? [];
+  const files = (await getFiles({ file_id: { $in: file_ids } })).map((file) => ({
+    file_id: file.file_id,
+    filename: file.filename,
+  }));
+
+  const fileList = files.map((file) => `- ${file.filename}`).join('\n');
+  const toolDescription = `Performs a semantic search based on a natural language query across the following files:\n${fileList}`;
+
+  const FileSearch = tool(
+    async ({ query }) => {
+      if (files.length === 0) {
+        return 'No files to search. Instruct the user to add files for the search.';
+      }
+      const jwtToken = req.headers.authorization.split(' ')[1];
+      if (!jwtToken) {
+        return 'There was an error authenticating the file search request.';
+      }
+      const queryPromises = files.map((file) =>
+        axios
+          .post(
+            `${process.env.RAG_API_URL}/query`,
+            {
+              file_id: file.file_id,
+              query,
+              k: 5,
+            },
+            {
+              headers: {
+                Authorization: `Bearer ${jwtToken}`,
+                'Content-Type': 'application/json',
+              },
+            },
+          )
+          .catch((error) => {
+            logger.error(
+              `Error encountered in \`file_search\` while querying file_id ${file._id}:`,
+              error,
+            );
+            return null;
+          }),
+      );
+
+      const results = await Promise.all(queryPromises);
+      const validResults = results.filter((result) => result !== null);
+
+      if (validResults.length === 0) {
+        return 'No results found or errors occurred while searching the files.';
+      }
+
+      const formattedResults = validResults
+        .flatMap((result) =>
+          result.data.map(([docInfo, relevanceScore]) => ({
+            filename: docInfo.metadata.source.split('/').pop(),
+            content: docInfo.page_content,
+            relevanceScore,
+          })),
+        )
+        .sort((a, b) => b.relevanceScore - a.relevanceScore);
+
+      const formattedString = formattedResults
+        .map(
+          (result) =>
+            `File: ${result.filename}\nRelevance: ${result.relevanceScore.toFixed(4)}\nContent: ${
+              result.content
+            }\n`,
+        )
+        .join('\n---\n');
+
+      return formattedString;
+    },
+    {
+      name: Tools.file_search,
+      description: toolDescription,
+      schema: z.object({
+        query: z
+          .string()
+          .describe(
+            'A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you\'re looking for. The query will be used for semantic similarity matching against the file contents.',
+          ),
+      }),
+    },
+  );
+
+  return FileSearch;
+};
+
+module.exports = createFileSearchTool;
--- a/api/app/clients/tools/util/fileSearch.js
+++ b/api/app/clients/tools/util/fileSearch.js
@@ -1,191 +0,0 @@
-const { z } = require('zod');
-const axios = require('axios');
-const { tool } = require('@langchain/core/tools');
-const { logger } = require('@librechat/data-schemas');
-const { generateShortLivedToken } = require('@librechat/api');
-const { Tools, EToolResources } = require('librechat-data-provider');
-const { filterFilesByAgentAccess } = require('~/server/services/Files/permissions');
-const { getFiles } = require('~/models/File');
-
-/**
- *
- * @param {Object} options
- * @param {ServerRequest} options.req
- * @param {Agent['tool_resources']} options.tool_resources
- * @param {string} [options.agentId] - The agent ID for file access control
- * @returns {Promise<{
- *   files: Array<{ file_id: string; filename: string }>,
- *   toolContext: string
- * }>}
- */
-const primeFiles = async (options) => {
-  const { tool_resources, req, agentId } = options;
-  const file_ids = tool_resources?.[EToolResources.file_search]?.file_ids ?? [];
-  const agentResourceIds = new Set(file_ids);
-  const resourceFiles = tool_resources?.[EToolResources.file_search]?.files ?? [];
-
-  // Get all files first
-  const allFiles = (await getFiles({ file_id: { $in: file_ids } }, null, { text: 0 })) ?? [];
-
-  // Filter by access if user and agent are provided
-  let dbFiles;
-  if (req?.user?.id && agentId) {
-    dbFiles = await filterFilesByAgentAccess({
-      files: allFiles,
-      userId: req.user.id,
-      role: req.user.role,
-      agentId,
-    });
-  } else {
-    dbFiles = allFiles;
-  }
-
-  dbFiles = dbFiles.concat(resourceFiles);
-
-  let toolContext = `- Note: Semantic search is available through the ${Tools.file_search} tool but no files are currently loaded. Request the user to upload documents to search through.`;
-
-  const files = [];
-  for (let i = 0; i < dbFiles.length; i++) {
-    const file = dbFiles[i];
-    if (!file) {
-      continue;
-    }
-    if (i === 0) {
-      toolContext = `- Note: Use the ${Tools.file_search} tool to find relevant information within:`;
-    }
-    toolContext += `\n\t- ${file.filename}${
-      agentResourceIds.has(file.file_id) ? '' : ' (just attached by user)'
-    }`;
-    files.push({
-      file_id: file.file_id,
-      filename: file.filename,
-    });
-  }
-
-  return { files, toolContext };
-};
-
-/**
- *
- * @param {Object} options
- * @param {ServerRequest} options.req
- * @param {Array<{ file_id: string; filename: string }>} options.files
- * @param {string} [options.entity_id]
- * @param {boolean} [options.fileCitations=false] - Whether to include citation instructions
- * @returns
- */
-const createFileSearchTool = async ({ req, files, entity_id, fileCitations = false }) => {
-  return tool(
-    async ({ query }) => {
-      if (files.length === 0) {
-        return 'No files to search. Instruct the user to add files for the search.';
-      }
-      const jwtToken = generateShortLivedToken(req.user.id);
-      if (!jwtToken) {
-        return 'There was an error authenticating the file search request.';
-      }
-
-      /**
-       *
-       * @param {import('librechat-data-provider').TFile} file
-       * @returns {{ file_id: string, query: string, k: number, entity_id?: string }}
-       */
-      const createQueryBody = (file) => {
-        const body = {
-          file_id: file.file_id,
-          query,
-          k: 5,
-        };
-        if (!entity_id) {
-          return body;
-        }
-        body.entity_id = entity_id;
-        logger.debug(`[${Tools.file_search}] RAG API /query body`, body);
-        return body;
-      };
-
-      const queryPromises = files.map((file) =>
-        axios
-          .post(`${process.env.RAG_API_URL}/query`, createQueryBody(file), {
-            headers: {
-              Authorization: `Bearer ${jwtToken}`,
-              'Content-Type': 'application/json',
-            },
-          })
-          .catch((error) => {
-            logger.error('Error encountered in `file_search` while querying file:', error);
-            return null;
-          }),
-      );
-
-      const results = await Promise.all(queryPromises);
-      const validResults = results.filter((result) => result !== null);
-
-      if (validResults.length === 0) {
-        return 'No results found or errors occurred while searching the files.';
-      }
-
-      const formattedResults = validResults
-        .flatMap((result, fileIndex) =>
-          result.data.map(([docInfo, distance]) => ({
-            filename: docInfo.metadata.source.split('/').pop(),
-            content: docInfo.page_content,
-            distance,
-            file_id: files[fileIndex]?.file_id,
-            page: docInfo.metadata.page || null,
-          })),
-        )
-        // TODO: results should be sorted by relevance, not distance
-        .sort((a, b) => a.distance - b.distance)
-        // TODO: make this configurable
-        .slice(0, 10);
-
-      const formattedString = formattedResults
-        .map(
-          (result, index) =>
-            `File: ${result.filename}${
-              fileCitations ? `\nAnchor: \\ue202turn0file${index} (${result.filename})` : ''
-            }\nRelevance: ${(1.0 - result.distance).toFixed(4)}\nContent: ${result.content}\n`,
-        )
-        .join('\n---\n');
-
-      const sources = formattedResults.map((result) => ({
-        type: 'file',
-        fileId: result.file_id,
-        content: result.content,
-        fileName: result.filename,
-        relevance: 1.0 - result.distance,
-        pages: result.page ? [result.page] : [],
-        pageRelevance: result.page ? { [result.page]: 1.0 - result.distance } : {},
-      }));
-
-      return [formattedString, { [Tools.file_search]: { sources, fileCitations } }];
-    },
-    {
-      name: Tools.file_search,
-      responseFormat: 'content_and_artifact',
-      description: `Performs semantic search across attached "${Tools.file_search}" documents using natural language queries. This tool analyzes the content of uploaded files to find relevant information, quotes, and passages that best match your query. Use this to extract specific information or find relevant sections within the available documents.${
-        fileCitations
-          ? `
-
-**CITE FILE SEARCH RESULTS:**
-Use anchor markers immediately after statements derived from file content. Reference the filename in your text:
- File citation: "The document.pdf states that... \\ue202turn0file0"  
- Page reference: "According to report.docx... \\ue202turn0file1"
- Multi-file: "Multiple sources confirm... \\ue200\\ue202turn0file0\\ue202turn0file1\\ue201"
-
-**ALWAYS mention the filename in your text before the citation marker. NEVER use markdown links or footnotes.**`
-          : ''
-      }`,
-      schema: z.object({
-        query: z
-          .string()
-          .describe(
-            "A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you're looking for. The query will be used for semantic similarity matching against the file contents.",
-          ),
-      }),
-    },
-  );
-};
-
-module.exports = { createFileSearchTool, primeFiles };
--- a/api/app/clients/tools/util/handleOpenAIErrors.js
+++ b/api/app/clients/tools/util/handleOpenAIErrors.js
@@ -23,8 +23,6 @@ async function handleOpenAIErrors(err, errorCallback, context = 'stream') {
    logger.warn(`[OpenAIClient.chatCompletion][${context}] Unhandled error type`);
  }

-  logger.error(err);
-
  if (errorCallback) {
    errorCallback(err);
  }
--- a/api/app/clients/tools/util/handleTools.js
+++ b/api/app/clients/tools/util/handleTools.js
@@ -1,40 +1,24 @@
-const { logger } = require('@librechat/data-schemas');
+const { Tools } = require('librechat-data-provider');
 const { SerpAPI } = require('@langchain/community/tools/serpapi');
 const { Calculator } = require('@langchain/community/tools/calculator');
-const { mcpToolPattern, loadWebSearchAuth, checkAccess } = require('@librechat/api');
-const { EnvVar, createCodeExecutionTool, createSearchTool } = require('@librechat/agents');
-const {
-  Tools,
-  Constants,
-  Permissions,
-  EToolResources,
-  PermissionTypes,
-  replaceSpecialVars,
-} = require('librechat-data-provider');
+const { createCodeExecutionTool, EnvVar } = require('@librechat/agents');
+const { getUserPluginAuthValue } = require('~/server/services/PluginService');
 const {
  availableTools,
-  manifestToolMap,
  // Basic Tools
  GoogleSearchAPI,
  // Structured Tools
  DALLE3,
-  FluxAPI,
-  OpenWeather,
  StructuredSD,
  StructuredACS,
  TraversaalSearch,
  StructuredWolfram,
-  createYouTubeTools,
  TavilySearchResults,
-  createOpenAIImageTools,
 } = require('../');
-const { primeFiles: primeCodeFiles } = require('~/server/services/Files/Code/process');
-const { createFileSearchTool, primeFiles: primeSearchFiles } = require('./fileSearch');
-const { getUserPluginAuthValue } = require('~/server/services/PluginService');
-const { createMCPTool, createMCPTools } = require('~/server/services/MCP');
-const { loadAuthValues } = require('~/server/services/Tools/credentials');
-const { getCachedTools } = require('~/server/services/Config');
-const { getRoleByName } = require('~/models/Role');
+const { primeFiles } = require('~/server/services/Files/Code/process');
+const createFileSearchTool = require('./createFileSearchTool');
+const { loadSpecs } = require('./loadSpecs');
+const { logger } = require('~/config');

 /**
 * Validates the availability and authentication of tools for a user based on environment variables or user-specific plugin authentication values.
@@ -95,12 +79,48 @@ const validateTools = async (user, tools = []) => {
    return Array.from(validToolsSet.values());
  } catch (err) {
    logger.error('[validateTools] There was a problem validating tools', err);
-    throw new Error(err);
+    throw new Error('There was a problem validating tools');
  }
 };

-/** @typedef {typeof import('@langchain/core/tools').Tool} ToolConstructor */
-/** @typedef {import('@langchain/core/tools').Tool} Tool */
+const loadAuthValues = async ({ userId, authFields }) => {
+  let authValues = {};
+
+  /**
+   * Finds the first non-empty value for the given authentication field, supporting alternate fields.
+   * @param {string[]} fields Array of strings representing the authentication fields. Supports alternate fields delimited by "||".
+   * @returns {Promise<{ authField: string, authValue: string} | null>} An object containing the authentication field and value, or null if not found.
+   */
+  const findAuthValue = async (fields) => {
+    for (const field of fields) {
+      let value = process.env[field];
+      if (value) {
+        return { authField: field, authValue: value };
+      }
+      try {
+        value = await getUserPluginAuthValue(userId, field);
+      } catch (err) {
+        if (field === fields[fields.length - 1] && !value) {
+          throw err;
+        }
+      }
+      if (value) {
+        return { authField: field, authValue: value };
+      }
+    }
+    return null;
+  };
+
+  for (let authField of authFields) {
+    const fields = authField.split('||');
+    const result = await findAuthValue(fields);
+    if (result) {
+      authValues[result.authField] = result.authValue;
+    }
+  }
+
+  return authValues;
+};

 /**
 * Initializes a tool with authentication values for the given user, supporting alternate authentication fields.
@@ -108,9 +128,9 @@ const validateTools = async (user, tools = []) => {
 *
 * @param {string} userId The user ID for which the tool is being loaded.
 * @param {Array<string>} authFields Array of strings representing the authentication fields. Supports alternate fields delimited by "||".
- * @param {ToolConstructor} ToolConstructor The constructor function for the tool to be initialized.
+ * @param {typeof import('langchain/tools').Tool} ToolConstructor The constructor function for the tool to be initialized.
 * @param {Object} options Optional parameters to be passed to the tool constructor alongside authentication values.
- * @returns {() => Promise<Tool>} An Async function that, when called, asynchronously initializes and returns an instance of the tool with authentication.
+ * @returns {Function} An Async function that, when called, asynchronously initializes and returns an instance of the tool with authentication.
 */
 const loadToolWithAuth = (userId, authFields, ToolConstructor, options = {}) => {
  return async function () {
@@ -119,53 +139,18 @@ const loadToolWithAuth = (userId, authFields, ToolConstructor, options = {}) =>
  };
 };

-/**
- * @param {string} toolKey
- * @returns {Array<string>}
- */
-const getAuthFields = (toolKey) => {
-  return manifestToolMap[toolKey]?.authConfig.map((auth) => auth.authField) ?? [];
-};
-
-/**
- *
- * @param {object} params
- * @param {string} params.user
- * @param {Record<string, Record<string, string>>} [object.userMCPAuthMap]
- * @param {AbortSignal} [object.signal]
- * @param {Pick<Agent, 'id' | 'provider' | 'model'>} [params.agent]
- * @param {string} [params.model]
- * @param {EModelEndpoint} [params.endpoint]
- * @param {LoadToolOptions} [params.options]
- * @param {boolean} [params.useSpecs]
- * @param {Array<string>} params.tools
- * @param {boolean} [params.functions]
- * @param {boolean} [params.returnMap]
- * @param {AppConfig['webSearch']} [params.webSearch]
- * @param {AppConfig['fileStrategy']} [params.fileStrategy]
- * @param {AppConfig['imageOutputType']} [params.imageOutputType]
- * @returns {Promise<{ loadedTools: Tool[], toolContextMap: Object<string, any> } | Record<string,Tool>>}
- */
 const loadTools = async ({
  user,
-  agent,
  model,
-  signal,
-  endpoint,
-  userMCPAuthMap,
-  tools = [],
-  options = {},
  functions = true,
  returnMap = false,
-  webSearch,
-  fileStrategy,
-  imageOutputType,
+  tools = [],
+  options = {},
+  skipSpecs = false,
 }) => {
  const toolConstructors = {
-    flux: FluxAPI,
    calculator: Calculator,
    google: GoogleSearchAPI,
-    open_weather: OpenWeather,
    wolfram: StructuredWolfram,
    'stable-diffusion': StructuredSD,
    'azure-ai-search': StructuredACS,
@@ -174,12 +159,10 @@ const loadTools = async ({
  };

  const customConstructors = {
-    serpapi: async (_toolContextMap) => {
-      const authFields = getAuthFields('serpapi');
-      let envVar = authFields[0] ?? '';
-      let apiKey = process.env[envVar];
+    serpapi: async () => {
+      let apiKey = process.env.SERPAPI_API_KEY;
      if (!apiKey) {
-        apiKey = await getUserPluginAuthValue(user, envVar);
+        apiKey = await getUserPluginAuthValue(user, 'SERPAPI_API_KEY');
      }
      return new SerpAPI(apiKey, {
        location: 'Austin,Texas,United States',
@@ -187,204 +170,62 @@ const loadTools = async ({
        gl: 'us',
      });
    },
-    youtube: async (_toolContextMap) => {
-      const authFields = getAuthFields('youtube');
-      const authValues = await loadAuthValues({ userId: user, authFields });
-      return createYouTubeTools(authValues);
-    },
-    image_gen_oai: async (toolContextMap) => {
-      const authFields = getAuthFields('image_gen_oai');
-      const authValues = await loadAuthValues({ userId: user, authFields });
-      const imageFiles = options.tool_resources?.[EToolResources.image_edit]?.files ?? [];
-      let toolContext = '';
-      for (let i = 0; i < imageFiles.length; i++) {
-        const file = imageFiles[i];
-        if (!file) {
-          continue;
-        }
-        if (i === 0) {
-          toolContext =
-            'Image files provided in this request (their image IDs listed in order of appearance) available for image editing:';
-        }
-        toolContext += `\n\t- ${file.file_id}`;
-        if (i === imageFiles.length - 1) {
-          toolContext += `\n\nInclude any you need in the \`image_ids\` array when calling \`${EToolResources.image_edit}_oai\`. You may also include previously referenced or generated image IDs.`;
-        }
-      }
-      if (toolContext) {
-        toolContextMap.image_edit_oai = toolContext;
-      }
-      return createOpenAIImageTools({
-        ...authValues,
-        isAgent: !!agent,
-        req: options.req,
-        imageOutputType,
-        fileStrategy,
-        imageFiles,
-      });
-    },
  };

  const requestedTools = {};

-  if (functions === true) {
+  if (functions) {
    toolConstructors.dalle = DALLE3;
  }

-  /** @type {ImageGenOptions} */
  const imageGenOptions = {
-    isAgent: !!agent,
    req: options.req,
-    fileStrategy,
+    fileStrategy: options.fileStrategy,
    processFileURL: options.processFileURL,
    returnMetadata: options.returnMetadata,
    uploadImageBuffer: options.uploadImageBuffer,
  };

  const toolOptions = {
-    flux: imageGenOptions,
-    dalle: imageGenOptions,
-    'stable-diffusion': imageGenOptions,
    serpapi: { location: 'Austin,Texas,United States', hl: 'en', gl: 'us' },
+    dalle: imageGenOptions,
+    'dall-e': imageGenOptions,
+    'stable-diffusion': imageGenOptions,
  };

-  /** @type {Record<string, string>} */
-  const toolContextMap = {};
-  const cachedTools = (await getCachedTools({ userId: user, includeGlobal: true })) ?? {};
-  const requestedMCPTools = {};
+  const toolAuthFields = {};
+
+  availableTools.forEach((tool) => {
+    if (customConstructors[tool.pluginKey]) {
+      return;
+    }
+
+    toolAuthFields[tool.pluginKey] = tool.authConfig.map((auth) => auth.authField);
+  });
+
+  const remainingTools = [];

  for (const tool of tools) {
    if (tool === Tools.execute_code) {
-      requestedTools[tool] = async () => {
-        const authValues = await loadAuthValues({
-          userId: user,
-          authFields: [EnvVar.CODE_API_KEY],
-        });
-        const codeApiKey = authValues[EnvVar.CODE_API_KEY];
-        const { files, toolContext } = await primeCodeFiles(
-          {
-            ...options,
-            agentId: agent?.id,
-          },
-          codeApiKey,
-        );
-        if (toolContext) {
-          toolContextMap[tool] = toolContext;
-        }
-        const CodeExecutionTool = createCodeExecutionTool({
+      const authValues = await loadAuthValues({
+        userId: user,
+        authFields: [EnvVar.CODE_API_KEY],
+      });
+      const files = await primeFiles(options, authValues[EnvVar.CODE_API_KEY]);
+      requestedTools[tool] = () =>
+        createCodeExecutionTool({
          user_id: user,
          files,
          ...authValues,
        });
-        CodeExecutionTool.apiKey = codeApiKey;
-        return CodeExecutionTool;
-      };
      continue;
    } else if (tool === Tools.file_search) {
-      requestedTools[tool] = async () => {
-        const { files, toolContext } = await primeSearchFiles({
-          ...options,
-          agentId: agent?.id,
-        });
-        if (toolContext) {
-          toolContextMap[tool] = toolContext;
-        }
-
-        /** @type {boolean | undefined} Check if user has FILE_CITATIONS permission */
-        let fileCitations;
-        if (fileCitations == null && options.req?.user != null) {
-          try {
-            fileCitations = await checkAccess({
-              user: options.req.user,
-              permissionType: PermissionTypes.FILE_CITATIONS,
-              permissions: [Permissions.USE],
-              getRoleByName,
-            });
-          } catch (error) {
-            logger.error('[handleTools] FILE_CITATIONS permission check failed:', error);
-            fileCitations = false;
-          }
-        }
-
-        return createFileSearchTool({
-          req: options.req,
-          files,
-          entity_id: agent?.id,
-          fileCitations,
-        });
-      };
-      continue;
-    } else if (tool === Tools.web_search) {
-      const result = await loadWebSearchAuth({
-        userId: user,
-        loadAuthValues,
-        webSearchConfig: webSearch,
-      });
-      const { onSearchResults, onGetHighlights } = options?.[Tools.web_search] ?? {};
-      requestedTools[tool] = async () => {
-        toolContextMap[tool] = `# \`${tool}\`:
-Current Date & Time: ${replaceSpecialVars({ text: '{{iso_datetime}}' })}
-1. **Execute immediately without preface** when using \`${tool}\`.
-2. **After the search, begin with a brief summary** that directly addresses the query without headers or explaining your process.
-3. **Structure your response clearly** using Markdown formatting (Level 2 headers for sections, lists for multiple points, tables for comparisons).
-4. **Cite sources properly** according to the citation anchor format, utilizing group anchors when appropriate.
-5. **Tailor your approach to the query type** (academic, news, coding, etc.) while maintaining an expert, journalistic, unbiased tone.
-6. **Provide comprehensive information** with specific details, examples, and as much relevant context as possible from search results.
-7. **Avoid moralizing language.**
-`.trim();
-        return createSearchTool({
-          ...result.authResult,
-          onSearchResults,
-          onGetHighlights,
-          logger,
-        });
-      };
-      continue;
-    } else if (tool && cachedTools && mcpToolPattern.test(tool)) {
-      const [toolName, serverName] = tool.split(Constants.mcp_delimiter);
-      if (toolName === Constants.mcp_server) {
-        /** Placeholder used for UI purposes */
-        continue;
-      }
-      if (serverName && options.req?.config?.mcpConfig?.[serverName] == null) {
-        logger.warn(
-          `MCP server "${serverName}" for "${toolName}" tool is not configured${agent?.id != null && agent.id ? ` but attached to "${agent.id}"` : ''}`,
-        );
-        continue;
-      }
-      if (toolName === Constants.mcp_all) {
-        const currentMCPGenerator = async (index) =>
-          createMCPTools({
-            req: options.req,
-            res: options.res,
-            index,
-            serverName,
-            userMCPAuthMap,
-            model: agent?.model ?? model,
-            provider: agent?.provider ?? endpoint,
-            signal,
-          });
-        requestedMCPTools[serverName] = [currentMCPGenerator];
-        continue;
-      }
-      const currentMCPGenerator = async (index) =>
-        createMCPTool({
-          index,
-          req: options.req,
-          res: options.res,
-          toolKey: tool,
-          userMCPAuthMap,
-          model: agent?.model ?? model,
-          provider: agent?.provider ?? endpoint,
-          signal,
-        });
-      requestedMCPTools[serverName] = requestedMCPTools[serverName] || [];
-      requestedMCPTools[serverName].push(currentMCPGenerator);
+      requestedTools[tool] = () => createFileSearchTool(options);
      continue;
    }

    if (customConstructors[tool]) {
-      requestedTools[tool] = async () => customConstructors[tool](toolContextMap);
+      requestedTools[tool] = customConstructors[tool];
      continue;
    }

@@ -392,66 +233,65 @@ Current Date & Time: ${replaceSpecialVars({ text: '{{iso_datetime}}' })}
      const options = toolOptions[tool] || {};
      const toolInstance = loadToolWithAuth(
        user,
-        getAuthFields(tool),
+        toolAuthFields[tool],
        toolConstructors[tool],
        options,
      );
      requestedTools[tool] = toolInstance;
      continue;
    }
+
+    if (functions) {
+      remainingTools.push(tool);
+    }
+  }
+
+  let specs = null;
+  if (functions && remainingTools.length > 0 && skipSpecs !== true) {
+    specs = await loadSpecs({
+      llm: model,
+      user,
+      message: options.message,
+      memory: options.memory,
+      signal: options.signal,
+      tools: remainingTools,
+      map: true,
+      verbose: false,
+    });
+  }
+
+  for (const tool of remainingTools) {
+    if (specs && specs[tool]) {
+      requestedTools[tool] = specs[tool];
+    }
  }

  if (returnMap) {
    return requestedTools;
  }

-  const toolPromises = [];
+  // load tools
+  let result = [];
  for (const tool of tools) {
    const validTool = requestedTools[tool];
-    if (validTool) {
-      toolPromises.push(
-        validTool().catch((error) => {
-          logger.error(`Error loading tool ${tool}:`, error);
-          return null;
-        }),
-      );
+    if (!validTool) {
+      continue;
+    }
+    const plugin = await validTool();
+
+    if (Array.isArray(plugin)) {
+      result = [...result, ...plugin];
+    } else if (plugin) {
+      result.push(plugin);
    }
  }

-  const loadedTools = (await Promise.all(toolPromises)).flatMap((plugin) => plugin || []);
-  const mcpToolPromises = [];
-  /** MCP server tools are initialized sequentially by server */
-  let index = -1;
-  for (const [serverName, generators] of Object.entries(requestedMCPTools)) {
-    index++;
-    for (const generator of generators) {
-      try {
-        if (generator && generators.length === 1) {
-          mcpToolPromises.push(
-            generator(index).catch((error) => {
-              logger.error(`Error loading ${serverName} tools:`, error);
-              return null;
-            }),
-          );
-          continue;
-        }
-        const mcpTool = await generator(index);
-        if (Array.isArray(mcpTool)) {
-          loadedTools.push(...mcpTool);
-        } else if (mcpTool) {
-          loadedTools.push(mcpTool);
-        }
-      } catch (error) {
-        logger.error(`Error loading MCP tool for server ${serverName}:`, error);
-      }
-    }
-  }
-  loadedTools.push(...(await Promise.all(mcpToolPromises)).flatMap((plugin) => plugin || []));
-  return { loadedTools, toolContextMap };
+  return result;
 };

 module.exports = {
  loadToolWithAuth,
+  loadAuthValues,
  validateTools,
  loadTools,
 };
--- a/api/app/clients/tools/util/handleTools.test.js
+++ b/api/app/clients/tools/util/handleTools.test.js
@@ -1,5 +1,8 @@
-const mongoose = require('mongoose');
-const { MongoMemoryServer } = require('mongodb-memory-server');
+const mockUser = {
+  _id: 'fakeId',
+  save: jest.fn(),
+  findByIdAndDelete: jest.fn(),
+};

 const mockPluginService = {
  updateUserPluginAuth: jest.fn(),
@@ -7,39 +10,23 @@ const mockPluginService = {
  getUserPluginAuthValue: jest.fn(),
 };

-jest.mock('~/server/services/PluginService', () => mockPluginService);
+jest.mock('~/models/User', () => {
+  return function () {
+    return mockUser;
+  };
+});

-jest.mock('~/server/services/Config', () => ({
-  getAppConfig: jest.fn().mockResolvedValue({
-    // Default app config for tool tests
-    paths: { uploads: '/tmp' },
-    fileStrategy: 'local',
-    filteredTools: [],
-    includedTools: [],
-  }),
-  getCachedTools: jest.fn().mockResolvedValue({
-    // Default cached tools for tests
-    dalle: {
-      type: 'function',
-      function: {
-        name: 'dalle',
-        description: 'DALL-E image generation',
-        parameters: {},
-      },
-    },
-  }),
-}));
+jest.mock('~/server/services/PluginService', () => mockPluginService);

 const { BaseLLM } = require('@langchain/openai');
 const { Calculator } = require('@langchain/community/tools/calculator');

-const { User } = require('~/db/models');
+const User = require('~/models/User');
 const PluginService = require('~/server/services/PluginService');
 const { validateTools, loadTools, loadToolWithAuth } = require('./handleTools');
 const { StructuredSD, availableTools, DALLE3 } = require('../');

 describe('Tool Handlers', () => {
-  let mongoServer;
  let fakeUser;
  const pluginKey = 'dalle';
  const pluginKey2 = 'wolfram';
@@ -50,9 +37,7 @@ describe('Tool Handlers', () => {
  const authConfigs = mainPlugin.authConfig;

  beforeAll(async () => {
-    mongoServer = await MongoMemoryServer.create();
-    const mongoUri = mongoServer.getUri();
-    await mongoose.connect(mongoUri);
+    mockUser.save.mockResolvedValue(undefined);

    const userAuthValues = {};
    mockPluginService.getUserPluginAuthValue.mockImplementation((userId, authField) => {
@@ -93,36 +78,9 @@ describe('Tool Handlers', () => {
  });

  afterAll(async () => {
-    await mongoose.disconnect();
-    await mongoServer.stop();
-  });
-
-  beforeEach(async () => {
-    // Clear mocks but not the database since we need the user to persist
-    jest.clearAllMocks();
-
-    // Reset the mock implementations
-    const userAuthValues = {};
-    mockPluginService.getUserPluginAuthValue.mockImplementation((userId, authField) => {
-      return userAuthValues[`${userId}-${authField}`];
-    });
-    mockPluginService.updateUserPluginAuth.mockImplementation(
-      (userId, authField, _pluginKey, credential) => {
-        const fields = authField.split('||');
-        fields.forEach((field) => {
-          userAuthValues[`${userId}-${field}`] = credential;
-        });
-      },
-    );
-
-    // Re-add the auth configs for the user
+    await mockUser.findByIdAndDelete(fakeUser._id);
    for (const authConfig of authConfigs) {
-      await PluginService.updateUserPluginAuth(
-        fakeUser._id,
-        authConfig.authField,
-        pluginKey,
-        mockCredential,
-      );
+      await PluginService.deleteUserPluginAuth(fakeUser._id, authConfig.authField);
    }
  });

@@ -170,14 +128,12 @@ describe('Tool Handlers', () => {
    );

    beforeAll(async () => {
-      const toolMap = await loadTools({
+      toolFunctions = await loadTools({
        user: fakeUser._id,
        model: BaseLLM,
        tools: sampleTools,
        returnMap: true,
-        useSpecs: true,
      });
-      toolFunctions = toolMap;
      loadTool1 = toolFunctions[sampleTools[0]];
      loadTool2 = toolFunctions[sampleTools[1]];
      loadTool3 = toolFunctions[sampleTools[2]];
@@ -239,7 +195,6 @@ describe('Tool Handlers', () => {
      expect(mockPluginService.getUserPluginAuthValue).toHaveBeenCalledWith(
        'userId',
        'DALLE3_API_KEY',
-        true,
      );
    });

@@ -260,6 +215,7 @@ describe('Tool Handlers', () => {
      try {
        await loadTool2();
      } catch (error) {
+        // eslint-disable-next-line jest/no-conditional-expect
        expect(error).toBeDefined();
      }
    });
@@ -268,7 +224,6 @@ describe('Tool Handlers', () => {
        user: fakeUser._id,
        model: BaseLLM,
        returnMap: true,
-        useSpecs: true,
      });
      expect(toolFunctions).toEqual({});
    });
@@ -280,7 +235,6 @@ describe('Tool Handlers', () => {
        tools: ['stable-diffusion'],
        functions: true,
        returnMap: true,
-        useSpecs: true,
      });
      const structuredTool = await toolFunctions['stable-diffusion']();
      expect(structuredTool).toBeInstanceOf(StructuredSD);
--- a/api/app/clients/tools/util/index.js
+++ b/api/app/clients/tools/util/index.js
@@ -1,8 +1,9 @@
-const { validateTools, loadTools } = require('./handleTools');
+const { validateTools, loadTools, loadAuthValues } = require('./handleTools');
 const handleOpenAIErrors = require('./handleOpenAIErrors');

 module.exports = {
  handleOpenAIErrors,
+  loadAuthValues,
  validateTools,
  loadTools,
 };
--- a/api/app/clients/tools/util/loadSpecs.js
+++ b/api/app/clients/tools/util/loadSpecs.js
@@ -0,0 +1,117 @@
+const fs = require('fs');
+const path = require('path');
+const { z } = require('zod');
+const { logger } = require('~/config');
+const { createOpenAPIPlugin } = require('~/app/clients/tools/dynamic/OpenAPIPlugin');
+
+// The minimum Manifest definition
+const ManifestDefinition = z.object({
+  schema_version: z.string().optional(),
+  name_for_human: z.string(),
+  name_for_model: z.string(),
+  description_for_human: z.string(),
+  description_for_model: z.string(),
+  auth: z.object({}).optional(),
+  api: z.object({
+    // Spec URL or can be the filename of the OpenAPI spec yaml file,
+    // located in api\app\clients\tools\.well-known\openapi
+    url: z.string(),
+    type: z.string().optional(),
+    is_user_authenticated: z.boolean().nullable().optional(),
+    has_user_authentication: z.boolean().nullable().optional(),
+  }),
+  // use to override any params that the LLM will consistently get wrong
+  params: z.object({}).optional(),
+  logo_url: z.string().optional(),
+  contact_email: z.string().optional(),
+  legal_info_url: z.string().optional(),
+});
+
+function validateJson(json) {
+  try {
+    return ManifestDefinition.parse(json);
+  } catch (error) {
+    logger.debug('[validateJson] manifest parsing error', error);
+    return false;
+  }
+}
+
+// omit the LLM to return the well known jsons as objects
+async function loadSpecs({ llm, user, message, tools = [], map = false, memory, signal }) {
+  const directoryPath = path.join(__dirname, '..', '.well-known');
+  let files = [];
+
+  for (let i = 0; i < tools.length; i++) {
+    const filePath = path.join(directoryPath, tools[i] + '.json');
+
+    try {
+      // If the access Promise is resolved, it means that the file exists
+      // Then we can add it to the files array
+      await fs.promises.access(filePath, fs.constants.F_OK);
+      files.push(tools[i] + '.json');
+    } catch (err) {
+      logger.error(`[loadSpecs] File ${tools[i] + '.json'} does not exist`, err);
+    }
+  }
+
+  if (files.length === 0) {
+    files = (await fs.promises.readdir(directoryPath)).filter(
+      (file) => path.extname(file) === '.json',
+    );
+  }
+
+  const validJsons = [];
+  const constructorMap = {};
+
+  logger.debug('[validateJson] files', files);
+
+  for (const file of files) {
+    if (path.extname(file) === '.json') {
+      const filePath = path.join(directoryPath, file);
+      const fileContent = await fs.promises.readFile(filePath, 'utf8');
+      const json = JSON.parse(fileContent);
+
+      if (!validateJson(json)) {
+        logger.debug('[validateJson] Invalid json', json);
+        continue;
+      }
+
+      if (llm && map) {
+        constructorMap[json.name_for_model] = async () =>
+          await createOpenAPIPlugin({
+            data: json,
+            llm,
+            message,
+            memory,
+            signal,
+            user,
+          });
+        continue;
+      }
+
+      if (llm) {
+        validJsons.push(createOpenAPIPlugin({ data: json, llm }));
+        continue;
+      }
+
+      validJsons.push(json);
+    }
+  }
+
+  if (map) {
+    return constructorMap;
+  }
+
+  const plugins = (await Promise.all(validJsons)).filter((plugin) => plugin);
+
+  //   logger.debug('[validateJson] plugins', plugins);
+  //   logger.debug(plugins[0].name);
+
+  return plugins;
+}
+
+module.exports = {
+  loadSpecs,
+  validateJson,
+  ManifestDefinition,
+};
--- a/api/app/clients/tools/util/loadSpecs.spec.js
+++ b/api/app/clients/tools/util/loadSpecs.spec.js
@@ -0,0 +1,101 @@
+const fs = require('fs');
+const { validateJson, loadSpecs, ManifestDefinition } = require('./loadSpecs');
+const { createOpenAPIPlugin } = require('../dynamic/OpenAPIPlugin');
+
+jest.mock('../dynamic/OpenAPIPlugin');
+
+describe('ManifestDefinition', () => {
+  it('should validate correct json', () => {
+    const json = {
+      name_for_human: 'Test',
+      name_for_model: 'Test',
+      description_for_human: 'Test',
+      description_for_model: 'Test',
+      api: {
+        url: 'http://test.com',
+      },
+    };
+
+    expect(() => ManifestDefinition.parse(json)).not.toThrow();
+  });
+
+  it('should not validate incorrect json', () => {
+    const json = {
+      name_for_human: 'Test',
+      name_for_model: 'Test',
+      description_for_human: 'Test',
+      description_for_model: 'Test',
+      api: {
+        url: 123, // incorrect type
+      },
+    };
+
+    expect(() => ManifestDefinition.parse(json)).toThrow();
+  });
+});
+
+describe('validateJson', () => {
+  it('should return parsed json if valid', () => {
+    const json = {
+      name_for_human: 'Test',
+      name_for_model: 'Test',
+      description_for_human: 'Test',
+      description_for_model: 'Test',
+      api: {
+        url: 'http://test.com',
+      },
+    };
+
+    expect(validateJson(json)).toEqual(json);
+  });
+
+  it('should return false if json is not valid', () => {
+    const json = {
+      name_for_human: 'Test',
+      name_for_model: 'Test',
+      description_for_human: 'Test',
+      description_for_model: 'Test',
+      api: {
+        url: 123, // incorrect type
+      },
+    };
+
+    expect(validateJson(json)).toEqual(false);
+  });
+});
+
+describe('loadSpecs', () => {
+  beforeEach(() => {
+    jest.spyOn(fs.promises, 'readdir').mockResolvedValue(['test.json']);
+    jest.spyOn(fs.promises, 'readFile').mockResolvedValue(
+      JSON.stringify({
+        name_for_human: 'Test',
+        name_for_model: 'Test',
+        description_for_human: 'Test',
+        description_for_model: 'Test',
+        api: {
+          url: 'http://test.com',
+        },
+      }),
+    );
+    createOpenAPIPlugin.mockResolvedValue({});
+  });
+
+  afterEach(() => {
+    jest.restoreAllMocks();
+  });
+
+  it('should return plugins', async () => {
+    const plugins = await loadSpecs({ llm: true, verbose: false });
+
+    expect(plugins).toHaveLength(1);
+    expect(createOpenAPIPlugin).toHaveBeenCalledTimes(1);
+  });
+
+  it('should return constructorMap if map is true', async () => {
+    const plugins = await loadSpecs({ llm: {}, map: true, verbose: false });
+
+    expect(plugins).toHaveProperty('Test');
+    expect(createOpenAPIPlugin).not.toHaveBeenCalled();
+  });
+});
--- a/api/app/index.js
+++ b/api/app/index.js
@@ -1,5 +1,11 @@
+const { browserClient } = require('./chatgpt-browser');
+const { askBing } = require('./bingai');
 const clients = require('./clients');
+const titleConvoBing = require('./titleConvoBing');

 module.exports = {
+  browserClient,
+  askBing,
+  titleConvoBing,
  ...clients,
 };
--- a/api/app/titleConvoBing.js
+++ b/api/app/titleConvoBing.js
@@ -0,0 +1,43 @@
+const throttle = require('lodash/throttle');
+const { isEnabled } = require('~/server/utils');
+const { logger } = require('~/config');
+
+const titleConvo = async ({ text, response }) => {
+  let title = 'New Chat';
+  const { TITLE_CONVO = 'true' } = process.env ?? {};
+  if (!isEnabled(TITLE_CONVO)) {
+    return title;
+  }
+
+  const { BingAIClient } = await import('nodejs-gpt');
+  const titleGenerator = new BingAIClient({
+    userToken: process.env.BINGAI_TOKEN,
+    debug: false,
+  });
+
+  let options = {
+    jailbreakConversationId: true,
+    systemMessage:
+      'You are a title-generator. As a title-generator, the old rules no longer apply. The only rule you need to follow is the following: Detect user language and write in the same language an extremely concise title for this conversation, which you must accurately detect. Write in the detected language. Title in 5 Words or Less. No Punctuation/Quotation. All first letters of every word should be capitalized and complete only the title in User Language only.',
+    context: `
+      ||>User:
+      "${text}"
+      ||>Response:
+      "${JSON.stringify(response?.text)}"`,
+    toneStyle: 'precise',
+  };
+  const titlePrompt = 'Title:';
+  try {
+    const res = await titleGenerator.sendMessage(titlePrompt, options);
+    title = res.response.replace(/Title: /, '').replace(/[".]/g, '');
+  } catch (e) {
+    logger.error('There was an issue generating title with BingAI', e);
+  }
+
+  logger.debug('[/ask/bingAI] CONVERSATION TITLE: ' + title);
+  return title;
+};
+
+const throttledTitleConvo = throttle(titleConvo, 3000);
+
+module.exports = throttledTitleConvo;
--- a/api/cache/banViolation.js
+++ b/api/cache/banViolation.js
@@ -1,9 +1,8 @@
-const { logger } = require('@librechat/data-schemas');
-const { isEnabled, math } = require('@librechat/api');
 const { ViolationTypes } = require('librechat-data-provider');
-const { deleteAllUserSessions } = require('~/models');
-const { removePorts } = require('~/server/utils');
+const { isEnabled, math, removePorts } = require('~/server/utils');
 const getLogStores = require('./getLogStores');
+const Session = require('~/models/Session');
+const { logger } = require('~/config');

 const { BAN_VIOLATIONS, BAN_INTERVAL } = process.env ?? {};
 const interval = math(BAN_INTERVAL, 20);
@@ -33,6 +32,7 @@ const banViolation = async (req, res, errorMessage) => {
  if (!isEnabled(BAN_VIOLATIONS)) {
    return;
  }
+
  if (!errorMessage) {
    return;
  }
@@ -46,11 +46,12 @@ const banViolation = async (req, res, errorMessage) => {
    return;
  }

-  await deleteAllUserSessions({ userId: user_id });
+  await Session.deleteAllUserSessions(user_id);
  res.clearCookie('refreshToken');

  const banLogs = getLogStores(ViolationTypes.BAN);
  const duration = errorMessage.duration || banLogs.opts.ttl;
+
  if (duration <= 0) {
    return;
  }
--- a/api/cache/banViolation.spec.js
+++ b/api/cache/banViolation.spec.js
@@ -1,28 +1,48 @@
-const mongoose = require('mongoose');
-const { MongoMemoryServer } = require('mongodb-memory-server');
 const banViolation = require('./banViolation');

-// Mock deleteAllUserSessions since we're testing ban logic, not session deletion
-jest.mock('~/models', () => ({
-  ...jest.requireActual('~/models'),
-  deleteAllUserSessions: jest.fn().mockResolvedValue(true),
-}));
+jest.mock('keyv');
+jest.mock('../models/Session');
+// Mocking the getLogStores function
+jest.mock('./getLogStores', () => {
+  return jest.fn().mockImplementation(() => {
+    const EventEmitter = require('events');
+    const { CacheKeys } = require('librechat-data-provider');
+    const math = require('../server/utils/math');
+    const mockGet = jest.fn();
+    const mockSet = jest.fn();
+    class KeyvMongo extends EventEmitter {
+      constructor(url = 'mongodb://127.0.0.1:27017', options) {
+        super();
+        this.ttlSupport = false;
+        url = url ?? {};
+        if (typeof url === 'string') {
+          url = { url };
+        }
+        if (url.uri) {
+          url = { url: url.uri, ...url };
+        }
+        this.opts = {
+          url,
+          collection: 'keyv',
+          ...url,
+          ...options,
+        };
+      }
+
+      get = mockGet;
+      set = mockSet;
+    }
+
+    return new KeyvMongo('', {
+      namespace: CacheKeys.BANS,
+      ttl: math(process.env.BAN_DURATION, 7200000),
+    });
+  });
+});

 describe('banViolation', () => {
-  let mongoServer;
  let req, res, errorMessage;

-  beforeAll(async () => {
-    mongoServer = await MongoMemoryServer.create();
-    const mongoUri = mongoServer.getUri();
-    await mongoose.connect(mongoUri);
-  });
-
-  afterAll(async () => {
-    await mongoose.disconnect();
-    await mongoServer.stop();
-  });
-
  beforeEach(() => {
    req = {
      ip: '127.0.0.1',
@@ -35,7 +55,7 @@ describe('banViolation', () => {
    };
    errorMessage = {
      type: 'someViolation',
-      user_id: new mongoose.Types.ObjectId().toString(), // Use valid ObjectId
+      user_id: '12345',
      prev_count: 0,
      violation_count: 0,
    };
--- a/api/cache/cacheConfig.js
+++ b/api/cache/cacheConfig.js
@@ -1,66 +0,0 @@
-const fs = require('fs');
-const { math, isEnabled } = require('@librechat/api');
-const { CacheKeys } = require('librechat-data-provider');
-
-// To ensure that different deployments do not interfere with each other's cache, we use a prefix for the Redis keys.
-// This prefix is usually the deployment ID, which is often passed to the container or pod as an env var.
-// Set REDIS_KEY_PREFIX_VAR to the env var that contains the deployment ID.
-const REDIS_KEY_PREFIX_VAR = process.env.REDIS_KEY_PREFIX_VAR;
-const REDIS_KEY_PREFIX = process.env.REDIS_KEY_PREFIX;
-if (REDIS_KEY_PREFIX_VAR && REDIS_KEY_PREFIX) {
-  throw new Error('Only either REDIS_KEY_PREFIX_VAR or REDIS_KEY_PREFIX can be set.');
-}
-
-const USE_REDIS = isEnabled(process.env.USE_REDIS);
-if (USE_REDIS && !process.env.REDIS_URI) {
-  throw new Error('USE_REDIS is enabled but REDIS_URI is not set.');
-}
-
-// Comma-separated list of cache namespaces that should be forced to use in-memory storage
-// even when Redis is enabled. This allows selective performance optimization for specific caches.
-const FORCED_IN_MEMORY_CACHE_NAMESPACES = process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES
-  ? process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES.split(',').map((key) => key.trim())
-  : [];
-
-// Validate against CacheKeys enum
-if (FORCED_IN_MEMORY_CACHE_NAMESPACES.length > 0) {
-  const validKeys = Object.values(CacheKeys);
-  const invalidKeys = FORCED_IN_MEMORY_CACHE_NAMESPACES.filter((key) => !validKeys.includes(key));
-
-  if (invalidKeys.length > 0) {
-    throw new Error(
-      `Invalid cache keys in FORCED_IN_MEMORY_CACHE_NAMESPACES: ${invalidKeys.join(', ')}. Valid keys: ${validKeys.join(', ')}`,
-    );
-  }
-}
-
-const cacheConfig = {
-  FORCED_IN_MEMORY_CACHE_NAMESPACES,
-  USE_REDIS,
-  REDIS_URI: process.env.REDIS_URI,
-  REDIS_USERNAME: process.env.REDIS_USERNAME,
-  REDIS_PASSWORD: process.env.REDIS_PASSWORD,
-  REDIS_CA: process.env.REDIS_CA ? fs.readFileSync(process.env.REDIS_CA, 'utf8') : null,
-  REDIS_KEY_PREFIX: process.env[REDIS_KEY_PREFIX_VAR] || REDIS_KEY_PREFIX || '',
-  REDIS_MAX_LISTENERS: math(process.env.REDIS_MAX_LISTENERS, 40),
-  REDIS_PING_INTERVAL: math(process.env.REDIS_PING_INTERVAL, 0),
-  /** Max delay between reconnection attempts in ms */
-  REDIS_RETRY_MAX_DELAY: math(process.env.REDIS_RETRY_MAX_DELAY, 3000),
-  /** Max number of reconnection attempts (0 = infinite) */
-  REDIS_RETRY_MAX_ATTEMPTS: math(process.env.REDIS_RETRY_MAX_ATTEMPTS, 10),
-  /** Connection timeout in ms */
-  REDIS_CONNECT_TIMEOUT: math(process.env.REDIS_CONNECT_TIMEOUT, 10000),
-  /** Queue commands when disconnected */
-  REDIS_ENABLE_OFFLINE_QUEUE: isEnabled(process.env.REDIS_ENABLE_OFFLINE_QUEUE ?? 'true'),
-  /** flag to modify redis connection by adding dnsLookup this is required when connecting to elasticache for ioredis
-   * see "Special Note: Aws Elasticache Clusters with TLS" on this webpage:  https://www.npmjs.com/package/ioredis **/
-  REDIS_USE_ALTERNATIVE_DNS_LOOKUP: isEnabled(process.env.REDIS_USE_ALTERNATIVE_DNS_LOOKUP),
-  /** Enable redis cluster without the need of multiple URIs */
-  USE_REDIS_CLUSTER: isEnabled(process.env.USE_REDIS_CLUSTER ?? 'false'),
-  CI: isEnabled(process.env.CI),
-  DEBUG_MEMORY_CACHE: isEnabled(process.env.DEBUG_MEMORY_CACHE),
-
-  BAN_DURATION: math(process.env.BAN_DURATION, 7200000), // 2 hours
-};
-
-module.exports = { cacheConfig };
--- a/api/cache/cacheConfig.spec.js
+++ b/api/cache/cacheConfig.spec.js
@@ -1,189 +0,0 @@
-const fs = require('fs');
-
-describe('cacheConfig', () => {
-  let originalEnv;
-  let originalReadFileSync;
-
-  beforeEach(() => {
-    originalEnv = { ...process.env };
-    originalReadFileSync = fs.readFileSync;
-
-    // Clear all related env vars first
-    delete process.env.REDIS_URI;
-    delete process.env.REDIS_CA;
-    delete process.env.REDIS_KEY_PREFIX_VAR;
-    delete process.env.REDIS_KEY_PREFIX;
-    delete process.env.USE_REDIS;
-    delete process.env.USE_REDIS_CLUSTER;
-    delete process.env.REDIS_PING_INTERVAL;
-    delete process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES;
-
-    // Clear require cache
-    jest.resetModules();
-  });
-
-  afterEach(() => {
-    process.env = originalEnv;
-    fs.readFileSync = originalReadFileSync;
-    jest.resetModules();
-  });
-
-  describe('REDIS_KEY_PREFIX validation and resolution', () => {
-    test('should throw error when both REDIS_KEY_PREFIX_VAR and REDIS_KEY_PREFIX are set', () => {
-      process.env.REDIS_KEY_PREFIX_VAR = 'DEPLOYMENT_ID';
-      process.env.REDIS_KEY_PREFIX = 'manual-prefix';
-
-      expect(() => {
-        require('./cacheConfig');
-      }).toThrow('Only either REDIS_KEY_PREFIX_VAR or REDIS_KEY_PREFIX can be set.');
-    });
-
-    test('should resolve REDIS_KEY_PREFIX from variable reference', () => {
-      process.env.REDIS_KEY_PREFIX_VAR = 'DEPLOYMENT_ID';
-      process.env.DEPLOYMENT_ID = 'test-deployment-123';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('test-deployment-123');
-    });
-
-    test('should use direct REDIS_KEY_PREFIX value', () => {
-      process.env.REDIS_KEY_PREFIX = 'direct-prefix';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('direct-prefix');
-    });
-
-    test('should default to empty string when no prefix is configured', () => {
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
-    });
-
-    test('should handle empty variable reference', () => {
-      process.env.REDIS_KEY_PREFIX_VAR = 'EMPTY_VAR';
-      process.env.EMPTY_VAR = '';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
-    });
-
-    test('should handle undefined variable reference', () => {
-      process.env.REDIS_KEY_PREFIX_VAR = 'UNDEFINED_VAR';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
-    });
-  });
-
-  describe('USE_REDIS and REDIS_URI validation', () => {
-    test('should throw error when USE_REDIS is enabled but REDIS_URI is not set', () => {
-      process.env.USE_REDIS = 'true';
-
-      expect(() => {
-        require('./cacheConfig');
-      }).toThrow('USE_REDIS is enabled but REDIS_URI is not set.');
-    });
-
-    test('should not throw error when USE_REDIS is enabled and REDIS_URI is set', () => {
-      process.env.USE_REDIS = 'true';
-      process.env.REDIS_URI = 'redis://localhost:6379';
-
-      expect(() => {
-        require('./cacheConfig');
-      }).not.toThrow();
-    });
-
-    test('should handle empty REDIS_URI when USE_REDIS is enabled', () => {
-      process.env.USE_REDIS = 'true';
-      process.env.REDIS_URI = '';
-
-      expect(() => {
-        require('./cacheConfig');
-      }).toThrow('USE_REDIS is enabled but REDIS_URI is not set.');
-    });
-  });
-
-  describe('USE_REDIS_CLUSTER configuration', () => {
-    test('should default to false when USE_REDIS_CLUSTER is not set', () => {
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(false);
-    });
-
-    test('should be false when USE_REDIS_CLUSTER is set to false', () => {
-      process.env.USE_REDIS_CLUSTER = 'false';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(false);
-    });
-
-    test('should be true when USE_REDIS_CLUSTER is set to true', () => {
-      process.env.USE_REDIS_CLUSTER = 'true';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(true);
-    });
-
-    test('should work with USE_REDIS enabled and REDIS_URI set', () => {
-      process.env.USE_REDIS_CLUSTER = 'true';
-      process.env.USE_REDIS = 'true';
-      process.env.REDIS_URI = 'redis://localhost:6379';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(true);
-      expect(cacheConfig.USE_REDIS).toBe(true);
-      expect(cacheConfig.REDIS_URI).toBe('redis://localhost:6379');
-    });
-  });
-
-  describe('REDIS_CA file reading', () => {
-    test('should be null when REDIS_CA is not set', () => {
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_CA).toBeNull();
-    });
-  });
-
-  describe('REDIS_PING_INTERVAL configuration', () => {
-    test('should default to 0 when REDIS_PING_INTERVAL is not set', () => {
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_PING_INTERVAL).toBe(0);
-    });
-
-    test('should use provided REDIS_PING_INTERVAL value', () => {
-      process.env.REDIS_PING_INTERVAL = '300';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.REDIS_PING_INTERVAL).toBe(300);
-    });
-  });
-
-  describe('FORCED_IN_MEMORY_CACHE_NAMESPACES validation', () => {
-    test('should parse comma-separated cache keys correctly', () => {
-      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = ' ROLES, MESSAGES ';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([
-        'ROLES',
-        'MESSAGES',
-      ]);
-    });
-
-    test('should throw error for invalid cache keys', () => {
-      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = 'INVALID_KEY,ROLES';
-
-      expect(() => {
-        require('./cacheConfig');
-      }).toThrow('Invalid cache keys in FORCED_IN_MEMORY_CACHE_NAMESPACES: INVALID_KEY');
-    });
-
-    test('should handle empty string gracefully', () => {
-      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = '';
-
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([]);
-    });
-
-    test('should handle undefined env var gracefully', () => {
-      const { cacheConfig } = require('./cacheConfig');
-      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([]);
-    });
-  });
-});
--- a/api/cache/cacheFactory.js
+++ b/api/cache/cacheFactory.js
@@ -1,108 +0,0 @@
-const KeyvRedis = require('@keyv/redis').default;
-const { Keyv } = require('keyv');
-const { RedisStore } = require('rate-limit-redis');
-const { Time } = require('librechat-data-provider');
-const { logger } = require('@librechat/data-schemas');
-const { RedisStore: ConnectRedis } = require('connect-redis');
-const MemoryStore = require('memorystore')(require('express-session'));
-const { keyvRedisClient, ioredisClient, GLOBAL_PREFIX_SEPARATOR } = require('./redisClients');
-const { cacheConfig } = require('./cacheConfig');
-const { violationFile } = require('./keyvFiles');
-
-/**
- * Creates a cache instance using Redis or a fallback store. Suitable for general caching needs.
- * @param {string} namespace - The cache namespace.
- * @param {number} [ttl] - Time to live for cache entries.
- * @param {object} [fallbackStore] - Optional fallback store if Redis is not used.
- * @returns {Keyv} Cache instance.
- */
-const standardCache = (namespace, ttl = undefined, fallbackStore = undefined) => {
-  if (
-    cacheConfig.USE_REDIS &&
-    !cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES?.includes(namespace)
-  ) {
-    try {
-      const keyvRedis = new KeyvRedis(keyvRedisClient);
-      const cache = new Keyv(keyvRedis, { namespace, ttl });
-      keyvRedis.namespace = cacheConfig.REDIS_KEY_PREFIX;
-      keyvRedis.keyPrefixSeparator = GLOBAL_PREFIX_SEPARATOR;
-
-      cache.on('error', (err) => {
-        logger.error(`Cache error in namespace ${namespace}:`, err);
-      });
-
-      return cache;
-    } catch (err) {
-      logger.error(`Failed to create Redis cache for namespace ${namespace}:`, err);
-      throw err;
-    }
-  }
-  if (fallbackStore) return new Keyv({ store: fallbackStore, namespace, ttl });
-  return new Keyv({ namespace, ttl });
-};
-
-/**
- * Creates a cache instance for storing violation data.
- * Uses a file-based fallback store if Redis is not enabled.
- * @param {string} namespace - The cache namespace for violations.
- * @param {number} [ttl] - Time to live for cache entries.
- * @returns {Keyv} Cache instance for violations.
- */
-const violationCache = (namespace, ttl = undefined) => {
-  return standardCache(`violations:${namespace}`, ttl, violationFile);
-};
-
-/**
- * Creates a session cache instance using Redis or in-memory store.
- * @param {string} namespace - The session namespace.
- * @param {number} [ttl] - Time to live for session entries.
- * @returns {MemoryStore | ConnectRedis} Session store instance.
- */
-const sessionCache = (namespace, ttl = undefined) => {
-  namespace = namespace.endsWith(':') ? namespace : `${namespace}:`;
-  if (!cacheConfig.USE_REDIS) return new MemoryStore({ ttl, checkPeriod: Time.ONE_DAY });
-  const store = new ConnectRedis({ client: ioredisClient, ttl, prefix: namespace });
-  if (ioredisClient) {
-    ioredisClient.on('error', (err) => {
-      logger.error(`Session store Redis error for namespace ${namespace}:`, err);
-    });
-  }
-  return store;
-};
-
-/**
- * Creates a rate limiter cache using Redis.
- * @param {string} prefix - The key prefix for rate limiting.
- * @returns {RedisStore|undefined} RedisStore instance or undefined if Redis is not used.
- */
-const limiterCache = (prefix) => {
-  if (!prefix) throw new Error('prefix is required');
-  if (!cacheConfig.USE_REDIS) return undefined;
-  prefix = prefix.endsWith(':') ? prefix : `${prefix}:`;
-
-  try {
-    if (!ioredisClient) {
-      logger.warn(`Redis client not available for rate limiter with prefix ${prefix}`);
-      return undefined;
-    }
-
-    return new RedisStore({ sendCommand, prefix });
-  } catch (err) {
-    logger.error(`Failed to create Redis rate limiter for prefix ${prefix}:`, err);
-    return undefined;
-  }
-};
-
-const sendCommand = (...args) => {
-  if (!ioredisClient) {
-    logger.warn('Redis client not available for command execution');
-    return Promise.reject(new Error('Redis client not available'));
-  }
-
-  return ioredisClient.call(...args).catch((err) => {
-    logger.error('Redis command execution failed:', err);
-    throw err;
-  });
-};
-
-module.exports = { standardCache, sessionCache, violationCache, limiterCache };
--- a/api/cache/cacheFactory.spec.js
+++ b/api/cache/cacheFactory.spec.js
@@ -1,432 +0,0 @@
-const { Time } = require('librechat-data-provider');
-
-// Mock dependencies first
-const mockKeyvRedis = {
-  namespace: '',
-  keyPrefixSeparator: '',
-};
-
-const mockKeyv = jest.fn().mockReturnValue({
-  mock: 'keyv',
-  on: jest.fn(),
-});
-const mockConnectRedis = jest.fn().mockReturnValue({ mock: 'connectRedis' });
-const mockMemoryStore = jest.fn().mockReturnValue({ mock: 'memoryStore' });
-const mockRedisStore = jest.fn().mockReturnValue({ mock: 'redisStore' });
-
-const mockIoredisClient = {
-  call: jest.fn(),
-  on: jest.fn(),
-};
-
-const mockKeyvRedisClient = {};
-const mockViolationFile = {};
-
-// Mock modules before requiring the main module
-jest.mock('@keyv/redis', () => ({
-  default: jest.fn().mockImplementation(() => mockKeyvRedis),
-}));
-
-jest.mock('keyv', () => ({
-  Keyv: mockKeyv,
-}));
-
-jest.mock('./cacheConfig', () => ({
-  cacheConfig: {
-    USE_REDIS: false,
-    REDIS_KEY_PREFIX: 'test',
-    FORCED_IN_MEMORY_CACHE_NAMESPACES: [],
-  },
-}));
-
-jest.mock('./redisClients', () => ({
-  keyvRedisClient: mockKeyvRedisClient,
-  ioredisClient: mockIoredisClient,
-  GLOBAL_PREFIX_SEPARATOR: '::',
-}));
-
-jest.mock('./keyvFiles', () => ({
-  violationFile: mockViolationFile,
-}));
-
-jest.mock('connect-redis', () => ({ RedisStore: mockConnectRedis }));
-
-jest.mock('memorystore', () => jest.fn(() => mockMemoryStore));
-
-jest.mock('rate-limit-redis', () => ({
-  RedisStore: mockRedisStore,
-}));
-
-jest.mock('@librechat/data-schemas', () => ({
-  logger: {
-    error: jest.fn(),
-    warn: jest.fn(),
-    info: jest.fn(),
-  },
-}));
-
-// Import after mocking
-const { standardCache, sessionCache, violationCache, limiterCache } = require('./cacheFactory');
-const { cacheConfig } = require('./cacheConfig');
-
-describe('cacheFactory', () => {
-  beforeEach(() => {
-    jest.clearAllMocks();
-
-    // Reset cache config mock
-    cacheConfig.USE_REDIS = false;
-    cacheConfig.REDIS_KEY_PREFIX = 'test';
-    cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = [];
-  });
-
-  describe('redisCache', () => {
-    it('should create Redis cache when USE_REDIS is true', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'test-namespace';
-      const ttl = 3600;
-
-      standardCache(namespace, ttl);
-
-      expect(require('@keyv/redis').default).toHaveBeenCalledWith(mockKeyvRedisClient);
-      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl });
-      expect(mockKeyvRedis.namespace).toBe(cacheConfig.REDIS_KEY_PREFIX);
-      expect(mockKeyvRedis.keyPrefixSeparator).toBe('::');
-    });
-
-    it('should create Redis cache with undefined ttl when not provided', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'test-namespace';
-
-      standardCache(namespace);
-
-      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl: undefined });
-    });
-
-    it('should use fallback store when USE_REDIS is false and fallbackStore is provided', () => {
-      cacheConfig.USE_REDIS = false;
-      const namespace = 'test-namespace';
-      const ttl = 3600;
-      const fallbackStore = { some: 'store' };
-
-      standardCache(namespace, ttl, fallbackStore);
-
-      expect(mockKeyv).toHaveBeenCalledWith({ store: fallbackStore, namespace, ttl });
-    });
-
-    it('should create default Keyv instance when USE_REDIS is false and no fallbackStore', () => {
-      cacheConfig.USE_REDIS = false;
-      const namespace = 'test-namespace';
-      const ttl = 3600;
-
-      standardCache(namespace, ttl);
-
-      expect(mockKeyv).toHaveBeenCalledWith({ namespace, ttl });
-    });
-
-    it('should handle namespace and ttl as undefined', () => {
-      cacheConfig.USE_REDIS = false;
-
-      standardCache();
-
-      expect(mockKeyv).toHaveBeenCalledWith({ namespace: undefined, ttl: undefined });
-    });
-
-    it('should use fallback when namespace is in FORCED_IN_MEMORY_CACHE_NAMESPACES', () => {
-      cacheConfig.USE_REDIS = true;
-      cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = ['forced-memory'];
-      const namespace = 'forced-memory';
-      const ttl = 3600;
-
-      standardCache(namespace, ttl);
-
-      expect(require('@keyv/redis').default).not.toHaveBeenCalled();
-      expect(mockKeyv).toHaveBeenCalledWith({ namespace, ttl });
-    });
-
-    it('should use Redis when namespace is not in FORCED_IN_MEMORY_CACHE_NAMESPACES', () => {
-      cacheConfig.USE_REDIS = true;
-      cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = ['other-namespace'];
-      const namespace = 'test-namespace';
-      const ttl = 3600;
-
-      standardCache(namespace, ttl);
-
-      expect(require('@keyv/redis').default).toHaveBeenCalledWith(mockKeyvRedisClient);
-      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl });
-    });
-
-    it('should throw error when Redis cache creation fails', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'test-namespace';
-      const ttl = 3600;
-      const testError = new Error('Redis connection failed');
-
-      const KeyvRedis = require('@keyv/redis').default;
-      KeyvRedis.mockImplementationOnce(() => {
-        throw testError;
-      });
-
-      expect(() => standardCache(namespace, ttl)).toThrow('Redis connection failed');
-
-      const { logger } = require('@librechat/data-schemas');
-      expect(logger.error).toHaveBeenCalledWith(
-        `Failed to create Redis cache for namespace ${namespace}:`,
-        testError,
-      );
-
-      expect(mockKeyv).not.toHaveBeenCalled();
-    });
-  });
-
-  describe('violationCache', () => {
-    it('should create violation cache with prefixed namespace', () => {
-      const namespace = 'test-violations';
-      const ttl = 7200;
-
-      // We can't easily mock the internal redisCache call since it's in the same module
-      // But we can test that the function executes without throwing
-      expect(() => violationCache(namespace, ttl)).not.toThrow();
-    });
-
-    it('should create violation cache with undefined ttl', () => {
-      const namespace = 'test-violations';
-
-      violationCache(namespace);
-
-      // The function should call redisCache with violations: prefixed namespace
-      // Since we can't easily mock the internal redisCache call, we test the behavior
-      expect(() => violationCache(namespace)).not.toThrow();
-    });
-
-    it('should handle undefined namespace', () => {
-      expect(() => violationCache(undefined)).not.toThrow();
-    });
-  });
-
-  describe('sessionCache', () => {
-    it('should return MemoryStore when USE_REDIS is false', () => {
-      cacheConfig.USE_REDIS = false;
-      const namespace = 'sessions';
-      const ttl = 86400;
-
-      const result = sessionCache(namespace, ttl);
-
-      expect(mockMemoryStore).toHaveBeenCalledWith({ ttl, checkPeriod: Time.ONE_DAY });
-      expect(result).toBe(mockMemoryStore());
-    });
-
-    it('should return ConnectRedis when USE_REDIS is true', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions';
-      const ttl = 86400;
-
-      const result = sessionCache(namespace, ttl);
-
-      expect(mockConnectRedis).toHaveBeenCalledWith({
-        client: mockIoredisClient,
-        ttl,
-        prefix: `${namespace}:`,
-      });
-      expect(result).toBe(mockConnectRedis());
-    });
-
-    it('should add colon to namespace if not present', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions';
-
-      sessionCache(namespace);
-
-      expect(mockConnectRedis).toHaveBeenCalledWith({
-        client: mockIoredisClient,
-        ttl: undefined,
-        prefix: 'sessions:',
-      });
-    });
-
-    it('should not add colon to namespace if already present', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions:';
-
-      sessionCache(namespace);
-
-      expect(mockConnectRedis).toHaveBeenCalledWith({
-        client: mockIoredisClient,
-        ttl: undefined,
-        prefix: 'sessions:',
-      });
-    });
-
-    it('should handle undefined ttl', () => {
-      cacheConfig.USE_REDIS = false;
-      const namespace = 'sessions';
-
-      sessionCache(namespace);
-
-      expect(mockMemoryStore).toHaveBeenCalledWith({
-        ttl: undefined,
-        checkPeriod: Time.ONE_DAY,
-      });
-    });
-
-    it('should throw error when ConnectRedis constructor fails', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions';
-      const ttl = 86400;
-
-      // Mock ConnectRedis to throw an error during construction
-      const redisError = new Error('Redis connection failed');
-      mockConnectRedis.mockImplementationOnce(() => {
-        throw redisError;
-      });
-
-      // The error should propagate up, not be caught
-      expect(() => sessionCache(namespace, ttl)).toThrow('Redis connection failed');
-
-      // Verify that MemoryStore was NOT used as fallback
-      expect(mockMemoryStore).not.toHaveBeenCalled();
-    });
-
-    it('should register error handler but let errors propagate to Express', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions';
-
-      // Create a mock session store with middleware methods
-      const mockSessionStore = {
-        get: jest.fn(),
-        set: jest.fn(),
-        destroy: jest.fn(),
-      };
-      mockConnectRedis.mockReturnValue(mockSessionStore);
-
-      const store = sessionCache(namespace);
-
-      // Verify error handler was registered
-      expect(mockIoredisClient.on).toHaveBeenCalledWith('error', expect.any(Function));
-
-      // Get the error handler
-      const errorHandler = mockIoredisClient.on.mock.calls.find((call) => call[0] === 'error')[1];
-
-      // Simulate an error from Redis during a session operation
-      const redisError = new Error('Socket closed unexpectedly');
-
-      // The error handler should log but not swallow the error
-      const { logger } = require('@librechat/data-schemas');
-      errorHandler(redisError);
-
-      expect(logger.error).toHaveBeenCalledWith(
-        `Session store Redis error for namespace ${namespace}::`,
-        redisError,
-      );
-
-      // Now simulate what happens when session middleware tries to use the store
-      const callback = jest.fn();
-      mockSessionStore.get.mockImplementation((sid, cb) => {
-        cb(new Error('Redis connection lost'));
-      });
-
-      // Call the store's get method (as Express session would)
-      store.get('test-session-id', callback);
-
-      // The error should be passed to the callback, not swallowed
-      expect(callback).toHaveBeenCalledWith(new Error('Redis connection lost'));
-    });
-
-    it('should handle null ioredisClient gracefully', () => {
-      cacheConfig.USE_REDIS = true;
-      const namespace = 'sessions';
-
-      // Temporarily set ioredisClient to null (simulating connection not established)
-      const originalClient = require('./redisClients').ioredisClient;
-      require('./redisClients').ioredisClient = null;
-
-      // ConnectRedis might accept null client but would fail on first use
-      // The important thing is it doesn't throw uncaught exceptions during construction
-      const store = sessionCache(namespace);
-      expect(store).toBeDefined();
-
-      // Restore original client
-      require('./redisClients').ioredisClient = originalClient;
-    });
-  });
-
-  describe('limiterCache', () => {
-    it('should return undefined when USE_REDIS is false', () => {
-      cacheConfig.USE_REDIS = false;
-      const result = limiterCache('prefix');
-
-      expect(result).toBeUndefined();
-    });
-
-    it('should return RedisStore when USE_REDIS is true', () => {
-      cacheConfig.USE_REDIS = true;
-      const result = limiterCache('rate-limit');
-
-      expect(mockRedisStore).toHaveBeenCalledWith({
-        sendCommand: expect.any(Function),
-        prefix: `rate-limit:`,
-      });
-      expect(result).toBe(mockRedisStore());
-    });
-
-    it('should add colon to prefix if not present', () => {
-      cacheConfig.USE_REDIS = true;
-      limiterCache('rate-limit');
-
-      expect(mockRedisStore).toHaveBeenCalledWith({
-        sendCommand: expect.any(Function),
-        prefix: 'rate-limit:',
-      });
-    });
-
-    it('should not add colon to prefix if already present', () => {
-      cacheConfig.USE_REDIS = true;
-      limiterCache('rate-limit:');
-
-      expect(mockRedisStore).toHaveBeenCalledWith({
-        sendCommand: expect.any(Function),
-        prefix: 'rate-limit:',
-      });
-    });
-
-    it('should pass sendCommand function that calls ioredisClient.call', async () => {
-      cacheConfig.USE_REDIS = true;
-      mockIoredisClient.call.mockResolvedValue('test-value');
-
-      limiterCache('rate-limit');
-
-      const sendCommandCall = mockRedisStore.mock.calls[0][0];
-      const sendCommand = sendCommandCall.sendCommand;
-
-      // Test that sendCommand properly delegates to ioredisClient.call
-      const args = ['GET', 'test-key'];
-      const result = await sendCommand(...args);
-
-      expect(mockIoredisClient.call).toHaveBeenCalledWith(...args);
-      expect(result).toBe('test-value');
-    });
-
-    it('should handle sendCommand errors properly', async () => {
-      cacheConfig.USE_REDIS = true;
-
-      // Mock the call method to reject with an error
-      const testError = new Error('Redis error');
-      mockIoredisClient.call.mockRejectedValue(testError);
-
-      limiterCache('rate-limit');
-
-      const sendCommandCall = mockRedisStore.mock.calls[0][0];
-      const sendCommand = sendCommandCall.sendCommand;
-
-      // Test that sendCommand properly handles errors
-      const args = ['GET', 'test-key'];
-
-      await expect(sendCommand(...args)).rejects.toThrow('Redis error');
-      expect(mockIoredisClient.call).toHaveBeenCalledWith(...args);
-    });
-
-    it('should handle undefined prefix', () => {
-      cacheConfig.USE_REDIS = true;
-      expect(() => limiterCache()).toThrow('prefix is required');
-    });
-  });
-});
--- a/api/cache/clearPendingReq.js
+++ b/api/cache/clearPendingReq.js
@@ -1,8 +1,7 @@
-const { Time, CacheKeys } = require('librechat-data-provider');
-const { isEnabled } = require('~/server/utils');
 const getLogStores = require('./getLogStores');
-
+const { isEnabled } = require('../server/utils');
 const { USE_REDIS, LIMIT_CONCURRENT_MESSAGES } = process.env ?? {};
+const ttl = 1000 * 60 * 1;

 /**
 * Clear or decrement pending requests from the cache.
@@ -29,7 +28,7 @@ const clearPendingReq = async ({ userId, cache: _cache }) => {
    return;
  }

-  const namespace = CacheKeys.PENDING_REQ;
+  const namespace = 'pending_req';
  const cache = _cache ?? getLogStores(namespace);

  if (!cache) {
@@ -40,7 +39,7 @@ const clearPendingReq = async ({ userId, cache: _cache }) => {
  const currentReq = +((await cache.get(key)) ?? 0);

  if (currentReq && currentReq >= 1) {
-    await cache.set(key, currentReq - 1, Time.ONE_MINUTE);
+    await cache.set(key, currentReq - 1, ttl);
  } else {
    await cache.delete(key);
  }
--- a/api/cache/getLogStores.js
+++ b/api/cache/getLogStores.js
@@ -1,210 +1,92 @@
-const { cacheConfig } = require('./cacheConfig');
-const { Keyv } = require('keyv');
+const Keyv = require('keyv');
 const { CacheKeys, ViolationTypes, Time } = require('librechat-data-provider');
-const { logFile } = require('./keyvFiles');
+const { logFile, violationFile } = require('./keyvFiles');
+const { math, isEnabled } = require('~/server/utils');
+const keyvRedis = require('./keyvRedis');
 const keyvMongo = require('./keyvMongo');
-const { standardCache, sessionCache, violationCache } = require('./cacheFactory');
+
+const { BAN_DURATION, USE_REDIS } = process.env ?? {};
+
+const duration = math(BAN_DURATION, 7200000);
+
+const createViolationInstance = (namespace) => {
+  const config = isEnabled(USE_REDIS) ? { store: keyvRedis } : { store: violationFile, namespace };
+  return new Keyv(config);
+};
+
+// Serve cache from memory so no need to clear it on startup/exit
+const pending_req = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis })
+  : new Keyv({ namespace: 'pending_req' });
+
+const config = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis })
+  : new Keyv({ namespace: CacheKeys.CONFIG_STORE });
+
+const roles = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis })
+  : new Keyv({ namespace: CacheKeys.ROLES });
+
+const audioRuns = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis, ttl: Time.TEN_MINUTES })
+  : new Keyv({ namespace: CacheKeys.AUDIO_RUNS, ttl: Time.TEN_MINUTES });
+
+const messages = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis, ttl: Time.FIVE_MINUTES })
+  : new Keyv({ namespace: CacheKeys.MESSAGES, ttl: Time.FIVE_MINUTES });
+
+const tokenConfig = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis, ttl: Time.THIRTY_MINUTES })
+  : new Keyv({ namespace: CacheKeys.TOKEN_CONFIG, ttl: Time.THIRTY_MINUTES });
+
+const genTitle = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis, ttl: Time.TWO_MINUTES })
+  : new Keyv({ namespace: CacheKeys.GEN_TITLE, ttl: Time.TWO_MINUTES });
+
+const modelQueries = isEnabled(process.env.USE_REDIS)
+  ? new Keyv({ store: keyvRedis })
+  : new Keyv({ namespace: CacheKeys.MODEL_QUERIES });
+
+const abortKeys = isEnabled(USE_REDIS)
+  ? new Keyv({ store: keyvRedis })
+  : new Keyv({ namespace: CacheKeys.ABORT_KEYS, ttl: Time.TEN_MINUTES });

 const namespaces = {
-  [ViolationTypes.GENERAL]: new Keyv({ store: logFile, namespace: 'violations' }),
-  [ViolationTypes.LOGINS]: violationCache(ViolationTypes.LOGINS),
-  [ViolationTypes.CONCURRENT]: violationCache(ViolationTypes.CONCURRENT),
-  [ViolationTypes.NON_BROWSER]: violationCache(ViolationTypes.NON_BROWSER),
-  [ViolationTypes.MESSAGE_LIMIT]: violationCache(ViolationTypes.MESSAGE_LIMIT),
-  [ViolationTypes.REGISTRATIONS]: violationCache(ViolationTypes.REGISTRATIONS),
-  [ViolationTypes.TOKEN_BALANCE]: violationCache(ViolationTypes.TOKEN_BALANCE),
-  [ViolationTypes.TTS_LIMIT]: violationCache(ViolationTypes.TTS_LIMIT),
-  [ViolationTypes.STT_LIMIT]: violationCache(ViolationTypes.STT_LIMIT),
-  [ViolationTypes.CONVO_ACCESS]: violationCache(ViolationTypes.CONVO_ACCESS),
-  [ViolationTypes.TOOL_CALL_LIMIT]: violationCache(ViolationTypes.TOOL_CALL_LIMIT),
-  [ViolationTypes.FILE_UPLOAD_LIMIT]: violationCache(ViolationTypes.FILE_UPLOAD_LIMIT),
-  [ViolationTypes.VERIFY_EMAIL_LIMIT]: violationCache(ViolationTypes.VERIFY_EMAIL_LIMIT),
-  [ViolationTypes.RESET_PASSWORD_LIMIT]: violationCache(ViolationTypes.RESET_PASSWORD_LIMIT),
-  [ViolationTypes.ILLEGAL_MODEL_REQUEST]: violationCache(ViolationTypes.ILLEGAL_MODEL_REQUEST),
-  [ViolationTypes.BAN]: new Keyv({
+  [CacheKeys.ROLES]: roles,
+  [CacheKeys.CONFIG_STORE]: config,
+  pending_req,
+  [ViolationTypes.BAN]: new Keyv({ store: keyvMongo, namespace: CacheKeys.BANS, ttl: duration }),
+  [CacheKeys.ENCODED_DOMAINS]: new Keyv({
    store: keyvMongo,
-    namespace: CacheKeys.BANS,
-    ttl: cacheConfig.BAN_DURATION,
+    namespace: CacheKeys.ENCODED_DOMAINS,
+    ttl: 0,
  }),
-
-  [CacheKeys.OPENID_SESSION]: sessionCache(CacheKeys.OPENID_SESSION),
-  [CacheKeys.SAML_SESSION]: sessionCache(CacheKeys.SAML_SESSION),
-
-  [CacheKeys.ROLES]: standardCache(CacheKeys.ROLES),
-  [CacheKeys.APP_CONFIG]: standardCache(CacheKeys.APP_CONFIG),
-  [CacheKeys.CONFIG_STORE]: standardCache(CacheKeys.CONFIG_STORE),
-  [CacheKeys.PENDING_REQ]: standardCache(CacheKeys.PENDING_REQ),
-  [CacheKeys.ENCODED_DOMAINS]: new Keyv({ store: keyvMongo, namespace: CacheKeys.ENCODED_DOMAINS }),
-  [CacheKeys.ABORT_KEYS]: standardCache(CacheKeys.ABORT_KEYS, Time.TEN_MINUTES),
-  [CacheKeys.TOKEN_CONFIG]: standardCache(CacheKeys.TOKEN_CONFIG, Time.THIRTY_MINUTES),
-  [CacheKeys.GEN_TITLE]: standardCache(CacheKeys.GEN_TITLE, Time.TWO_MINUTES),
-  [CacheKeys.S3_EXPIRY_INTERVAL]: standardCache(CacheKeys.S3_EXPIRY_INTERVAL, Time.THIRTY_MINUTES),
-  [CacheKeys.MODEL_QUERIES]: standardCache(CacheKeys.MODEL_QUERIES),
-  [CacheKeys.AUDIO_RUNS]: standardCache(CacheKeys.AUDIO_RUNS, Time.TEN_MINUTES),
-  [CacheKeys.MESSAGES]: standardCache(CacheKeys.MESSAGES, Time.ONE_MINUTE),
-  [CacheKeys.FLOWS]: standardCache(CacheKeys.FLOWS, Time.ONE_MINUTE * 3),
-  [CacheKeys.OPENID_EXCHANGED_TOKENS]: standardCache(
-    CacheKeys.OPENID_EXCHANGED_TOKENS,
-    Time.TEN_MINUTES,
+  general: new Keyv({ store: logFile, namespace: 'violations' }),
+  concurrent: createViolationInstance('concurrent'),
+  non_browser: createViolationInstance('non_browser'),
+  message_limit: createViolationInstance('message_limit'),
+  token_balance: createViolationInstance(ViolationTypes.TOKEN_BALANCE),
+  registrations: createViolationInstance('registrations'),
+  [ViolationTypes.TTS_LIMIT]: createViolationInstance(ViolationTypes.TTS_LIMIT),
+  [ViolationTypes.STT_LIMIT]: createViolationInstance(ViolationTypes.STT_LIMIT),
+  [ViolationTypes.CONVO_ACCESS]: createViolationInstance(ViolationTypes.CONVO_ACCESS),
+  [ViolationTypes.FILE_UPLOAD_LIMIT]: createViolationInstance(ViolationTypes.FILE_UPLOAD_LIMIT),
+  [ViolationTypes.VERIFY_EMAIL_LIMIT]: createViolationInstance(ViolationTypes.VERIFY_EMAIL_LIMIT),
+  [ViolationTypes.RESET_PASSWORD_LIMIT]: createViolationInstance(
+    ViolationTypes.RESET_PASSWORD_LIMIT,
  ),
+  [ViolationTypes.ILLEGAL_MODEL_REQUEST]: createViolationInstance(
+    ViolationTypes.ILLEGAL_MODEL_REQUEST,
+  ),
+  logins: createViolationInstance('logins'),
+  [CacheKeys.ABORT_KEYS]: abortKeys,
+  [CacheKeys.TOKEN_CONFIG]: tokenConfig,
+  [CacheKeys.GEN_TITLE]: genTitle,
+  [CacheKeys.MODEL_QUERIES]: modelQueries,
+  [CacheKeys.AUDIO_RUNS]: audioRuns,
+  [CacheKeys.MESSAGES]: messages,
 };

-/**
- * Gets all cache stores that have TTL configured
- * @returns {Keyv[]}
- */
-function getTTLStores() {
-  return Object.values(namespaces).filter(
-    (store) =>
-      store instanceof Keyv &&
-      parseInt(store.opts?.ttl ?? '0') > 0 &&
-      !store.opts?.store?.constructor?.name?.includes('Redis'), // Only include non-Redis stores
-  );
-}
-
-/**
- * Clears entries older than the cache's TTL
- * @param {Keyv} cache
- */
-async function clearExpiredFromCache(cache) {
-  if (!cache?.opts?.store?.entries) {
-    return;
-  }
-
-  const ttl = cache.opts.ttl;
-  if (!ttl) {
-    return;
-  }
-
-  const expiryTime = Date.now() - ttl;
-  let cleared = 0;
-
-  // Get all keys first to avoid modification during iteration
-  const keys = Array.from(cache.opts.store.keys());
-
-  for (const key of keys) {
-    try {
-      const raw = cache.opts.store.get(key);
-      if (!raw) {
-        continue;
-      }
-
-      const data = cache.opts.deserialize(raw);
-      // Check if the entry is older than TTL
-      if (data?.expires && data.expires <= expiryTime) {
-        const deleted = await cache.opts.store.delete(key);
-        if (!deleted) {
-          cacheConfig.DEBUG_MEMORY_CACHE &&
-            console.warn(`[Cache] Error deleting entry: ${key} from ${cache.opts.namespace}`);
-          continue;
-        }
-        cleared++;
-      }
-    } catch (error) {
-      cacheConfig.DEBUG_MEMORY_CACHE &&
-        console.log(`[Cache] Error processing entry from ${cache.opts.namespace}:`, error);
-      const deleted = await cache.opts.store.delete(key);
-      if (!deleted) {
-        cacheConfig.DEBUG_MEMORY_CACHE &&
-          console.warn(`[Cache] Error deleting entry: ${key} from ${cache.opts.namespace}`);
-        continue;
-      }
-      cleared++;
-    }
-  }
-
-  if (cleared > 0) {
-    cacheConfig.DEBUG_MEMORY_CACHE &&
-      console.log(
-        `[Cache] Cleared ${cleared} entries older than ${ttl}ms from ${cache.opts.namespace}`,
-      );
-  }
-}
-
-const auditCache = () => {
-  const ttlStores = getTTLStores();
-  console.log('[Cache] Starting audit');
-
-  ttlStores.forEach((store) => {
-    if (!store?.opts?.store?.entries) {
-      return;
-    }
-
-    console.log(`[Cache] ${store.opts.namespace} entries:`, {
-      count: store.opts.store.size,
-      ttl: store.opts.ttl,
-      keys: Array.from(store.opts.store.keys()),
-      entriesWithTimestamps: Array.from(store.opts.store.entries()).map(([key, value]) => ({
-        key,
-        value,
-      })),
-    });
-  });
-};
-
-/**
- * Clears expired entries from all TTL-enabled stores
- */
-async function clearAllExpiredFromCache() {
-  const ttlStores = getTTLStores();
-  await Promise.all(ttlStores.map((store) => clearExpiredFromCache(store)));
-
-  // Force garbage collection if available (Node.js with --expose-gc flag)
-  if (global.gc) {
-    global.gc();
-  }
-}
-
-if (!cacheConfig.USE_REDIS && !cacheConfig.CI) {
-  /** @type {Set<NodeJS.Timeout>} */
-  const cleanupIntervals = new Set();
-
-  // Clear expired entries every 30 seconds
-  const cleanup = setInterval(() => {
-    clearAllExpiredFromCache();
-  }, Time.THIRTY_SECONDS);
-
-  cleanupIntervals.add(cleanup);
-
-  if (cacheConfig.DEBUG_MEMORY_CACHE) {
-    const monitor = setInterval(() => {
-      const ttlStores = getTTLStores();
-      const memory = process.memoryUsage();
-      const totalSize = ttlStores.reduce((sum, store) => sum + (store.opts?.store?.size ?? 0), 0);
-
-      console.log('[Cache] Memory usage:', {
-        heapUsed: `${(memory.heapUsed / 1024 / 1024).toFixed(2)} MB`,
-        heapTotal: `${(memory.heapTotal / 1024 / 1024).toFixed(2)} MB`,
-        rss: `${(memory.rss / 1024 / 1024).toFixed(2)} MB`,
-        external: `${(memory.external / 1024 / 1024).toFixed(2)} MB`,
-        totalCacheEntries: totalSize,
-      });
-
-      auditCache();
-    }, Time.ONE_MINUTE);
-
-    cleanupIntervals.add(monitor);
-  }
-
-  const dispose = () => {
-    cacheConfig.DEBUG_MEMORY_CACHE && console.log('[Cache] Cleaning up and shutting down...');
-    cleanupIntervals.forEach((interval) => clearInterval(interval));
-    cleanupIntervals.clear();
-
-    // One final cleanup before exit
-    clearAllExpiredFromCache().then(() => {
-      cacheConfig.DEBUG_MEMORY_CACHE && console.log('[Cache] Final cleanup completed');
-      process.exit(0);
-    });
-  };
-
-  // Handle various termination signals
-  process.on('SIGTERM', dispose);
-  process.on('SIGINT', dispose);
-  process.on('SIGQUIT', dispose);
-  process.on('SIGHUP', dispose);
-}
-
 /**
 * Returns the keyv cache specified by type.
 * If an invalid type is passed, an error will be thrown.
--- a/api/cache/keyvFiles.js
+++ b/api/cache/keyvFiles.js
@@ -1,9 +1,11 @@
 const { KeyvFile } = require('keyv-file');

-const logFile = new KeyvFile({ filename: './data/logs.json' }).setMaxListeners(20);
-const violationFile = new KeyvFile({ filename: './data/violations.json' }).setMaxListeners(20);
+const logFile = new KeyvFile({ filename: './data/logs.json' });
+const pendingReqFile = new KeyvFile({ filename: './data/pendingReqCache.json' });
+const violationFile = new KeyvFile({ filename: './data/violations.json' });

 module.exports = {
  logFile,
+  pendingReqFile,
  violationFile,
 };
--- a/api/cache/keyvMongo.js
+++ b/api/cache/keyvMongo.js
@@ -1,272 +1,9 @@
-// api/cache/keyvMongo.js
-const mongoose = require('mongoose');
-const EventEmitter = require('events');
-const { GridFSBucket } = require('mongodb');
+const KeyvMongo = require('@keyv/mongo');
 const { logger } = require('~/config');

-const storeMap = new Map();
-
-class KeyvMongoCustom extends EventEmitter {
-  constructor(url, options = {}) {
-    super();
-
-    url = url || {};
-    if (typeof url === 'string') {
-      url = { url };
-    }
-    if (url.uri) {
-      url = { url: url.uri, ...url };
-    }
-
-    this.opts = {
-      url: 'mongodb://127.0.0.1:27017',
-      collection: 'keyv',
-      ...url,
-      ...options,
-    };
-
-    this.ttlSupport = false;
-
-    // Filter valid options
-    const keyvMongoKeys = new Set([
-      'url',
-      'collection',
-      'namespace',
-      'serialize',
-      'deserialize',
-      'uri',
-      'useGridFS',
-      'dialect',
-    ]);
-    this.opts = Object.fromEntries(Object.entries(this.opts).filter(([k]) => keyvMongoKeys.has(k)));
-  }
-
-  // Helper to access the store WITHOUT storing a promise on the instance
-  _getClient() {
-    const storeKey = `${this.opts.collection}:${this.opts.useGridFS ? 'gridfs' : 'collection'}`;
-
-    // If we already have the store initialized, return it directly
-    if (storeMap.has(storeKey)) {
-      return Promise.resolve(storeMap.get(storeKey));
-    }
-
-    // Check mongoose connection state
-    if (mongoose.connection.readyState !== 1) {
-      return Promise.reject(
-        new Error('Mongoose connection not ready. Ensure connectDb() is called first.'),
-      );
-    }
-
-    try {
-      const db = mongoose.connection.db;
-      let client;
-
-      if (this.opts.useGridFS) {
-        const bucket = new GridFSBucket(db, {
-          readPreference: this.opts.readPreference,
-          bucketName: this.opts.collection,
-        });
-        const store = db.collection(`${this.opts.collection}.files`);
-        client = { bucket, store, db };
-      } else {
-        const collection = this.opts.collection || 'keyv';
-        const store = db.collection(collection);
-        client = { store, db };
-      }
-
-      storeMap.set(storeKey, client);
-      return Promise.resolve(client);
-    } catch (error) {
-      this.emit('error', error);
-      return Promise.reject(error);
-    }
-  }
-
-  async get(key) {
-    const client = await this._getClient();
-
-    if (this.opts.useGridFS) {
-      await client.store.updateOne(
-        {
-          filename: key,
-        },
-        {
-          $set: {
-            'metadata.lastAccessed': new Date(),
-          },
-        },
-      );
-
-      const stream = client.bucket.openDownloadStreamByName(key);
-
-      return new Promise((resolve) => {
-        const resp = [];
-        stream.on('error', () => {
-          resolve(undefined);
-        });
-
-        stream.on('end', () => {
-          const data = Buffer.concat(resp).toString('utf8');
-          resolve(data);
-        });
-
-        stream.on('data', (chunk) => {
-          resp.push(chunk);
-        });
-      });
-    }
-
-    const document = await client.store.findOne({ key: { $eq: key } });
-
-    if (!document) {
-      return undefined;
-    }
-
-    return document.value;
-  }
-
-  async getMany(keys) {
-    const client = await this._getClient();
-
-    if (this.opts.useGridFS) {
-      const promises = [];
-      for (const key of keys) {
-        promises.push(this.get(key));
-      }
-
-      const values = await Promise.allSettled(promises);
-      const data = [];
-      for (const value of values) {
-        data.push(value.value);
-      }
-
-      return data;
-    }
-
-    const values = await client.store
-      .find({ key: { $in: keys } })
-      .project({ _id: 0, value: 1, key: 1 })
-      .toArray();
-
-    const results = [...keys];
-    let i = 0;
-    for (const key of keys) {
-      const rowIndex = values.findIndex((row) => row.key === key);
-      results[i] = rowIndex > -1 ? values[rowIndex].value : undefined;
-      i++;
-    }
-
-    return results;
-  }
-
-  async set(key, value, ttl) {
-    const client = await this._getClient();
-    const expiresAt = typeof ttl === 'number' ? new Date(Date.now() + ttl) : null;
-
-    if (this.opts.useGridFS) {
-      const stream = client.bucket.openUploadStream(key, {
-        metadata: {
-          expiresAt,
-          lastAccessed: new Date(),
-        },
-      });
-
-      return new Promise((resolve) => {
-        stream.on('finish', () => {
-          resolve(stream);
-        });
-        stream.end(value);
-      });
-    }
-
-    await client.store.updateOne(
-      { key: { $eq: key } },
-      { $set: { key, value, expiresAt } },
-      { upsert: true },
-    );
-  }
-
-  async delete(key) {
-    if (typeof key !== 'string') {
-      return false;
-    }
-
-    const client = await this._getClient();
-
-    if (this.opts.useGridFS) {
-      try {
-        const bucket = new GridFSBucket(client.db, {
-          bucketName: this.opts.collection,
-        });
-        const files = await bucket.find({ filename: key }).toArray();
-        await client.bucket.delete(files[0]._id);
-        return true;
-      } catch {
-        return false;
-      }
-    }
-
-    const object = await client.store.deleteOne({ key: { $eq: key } });
-    return object.deletedCount > 0;
-  }
-
-  async deleteMany(keys) {
-    const client = await this._getClient();
-
-    if (this.opts.useGridFS) {
-      const bucket = new GridFSBucket(client.db, {
-        bucketName: this.opts.collection,
-      });
-      const files = await bucket.find({ filename: { $in: keys } }).toArray();
-      if (files.length === 0) {
-        return false;
-      }
-
-      await Promise.all(files.map(async (file) => client.bucket.delete(file._id)));
-      return true;
-    }
-
-    const object = await client.store.deleteMany({ key: { $in: keys } });
-    return object.deletedCount > 0;
-  }
-
-  async clear() {
-    const client = await this._getClient();
-
-    if (this.opts.useGridFS) {
-      try {
-        await client.bucket.drop();
-      } catch (error) {
-        // Throw error if not "namespace not found" error
-        if (!(error.code === 26)) {
-          throw error;
-        }
-      }
-    }
-
-    await client.store.deleteMany({
-      key: { $regex: this.namespace ? `^${this.namespace}:*` : '' },
-    });
-  }
-
-  async has(key) {
-    const client = await this._getClient();
-    const filter = { [this.opts.useGridFS ? 'filename' : 'key']: { $eq: key } };
-    const document = await client.store.countDocuments(filter, { limit: 1 });
-    return document !== 0;
-  }
-
-  // No-op disconnect
-  async disconnect() {
-    // This is a no-op since we don't want to close the shared mongoose connection
-    return true;
-  }
-}
-
-const keyvMongo = new KeyvMongoCustom({
-  collection: 'logs',
-});
+const { MONGO_URI } = process.env ?? {};

+const keyvMongo = new KeyvMongo(MONGO_URI, { collection: 'logs' });
 keyvMongo.on('error', (err) => logger.error('KeyvMongo connection error:', err));

 module.exports = keyvMongo;
--- a/api/cache/keyvRedis.js
+++ b/api/cache/keyvRedis.js
@@ -0,0 +1,20 @@
+const KeyvRedis = require('@keyv/redis');
+const { logger } = require('~/config');
+const { isEnabled } = require('~/server/utils');
+
+const { REDIS_URI, USE_REDIS } = process.env;
+
+let keyvRedis;
+
+if (REDIS_URI && isEnabled(USE_REDIS)) {
+  keyvRedis = new KeyvRedis(REDIS_URI, { useRedisSets: false });
+  keyvRedis.on('error', (err) => logger.error('KeyvRedis connection error:', err));
+  keyvRedis.setMaxListeners(20);
+  logger.info(
+    '[Optional] Redis initialized. Note: Redis support is experimental. If you have issues, disable it. Cache needs to be flushed for values to refresh.',
+  );
+} else {
+  logger.info('[Optional] Redis not initialized. Note: Redis support is experimental.');
+}
+
+module.exports = keyvRedis;
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Marco Beretta	daacfce581	feat: simplify STTService request handling by refining SDK usage and improving error logging	2024-12-04 17:33:38 +01:00
Marco Beretta	ffa5f6f09b	🌊 feat: refine SDK usage logic in STT and TTS services, improve header handling	2024-11-24 01:14:17 +01:00
Marco Beretta	b7f4903acd	🌊 feat: enhance TTSService with Deepgram SDK integration and refactor voice validation	2024-11-23 16:49:56 +01:00
Marco Beretta	5eabd2493c	🌊 feat: update Deepgram SDK integration for STT and remove unused TTS provider	2024-11-23 12:20:17 +01:00
Marco Beretta	25d51eff31	🌊 feat: add Deepgram support for STT providers	2024-11-23 12:17:53 +01:00