Skip to main content
LibreChat allows you to add custom AI endpoints beyond the built-in providers. Configure them in librechat.yaml under the endpoints.custom array.

​
Basic Structure

Each custom endpoint requires a unique name, API key, base URL, and models configuration. 
endpoints:
  custom:
    - name: 'endpoint-name'
      apiKey: '${ENV_VARIABLE}'
      baseURL: 'https://api.provider.com/v1'
      models:
        default: ['model-1', 'model-2']
        fetch: false

​
Configuration Fields

​
name
string
required
Unique name for the endpoint (displayed in the UI)
​
apiKey
string
required
API key for authentication (use environment variables)
apiKey: '${GROQ_API_KEY}'
​
baseURL
string
required
Base URL for the API endpoint
baseURL: 'https://api.groq.com/openai/v1/'

​
Models Configuration

​
models.default
array
required
List of default models to use (at least one required)
models:
  default:
    - 'llama3-70b-8192'
    - 'mixtral-8x7b-32768'
​
models.fetch
boolean
default:"false"
Fetch available models from API
When true, queries the API /models endpoint

​
Optional Settings

​
titleConvo
boolean
default:"false"
Enable automatic conversation title generation
​
titleModel
string
default:"gpt-3.5-turbo"
Model to use for generating titles
titleModel: 'mixtral-8x7b-32768'
​
titleMethod
string
default:"completion"
Method for title generation: completion or functions
​
summarize
boolean
default:"false"
Enable conversation summarization
​
summaryModel
string
default:"gpt-3.5-turbo"
Model to use for summarization
​
modelDisplayLabel
string
default:"AI"
Label displayed for AI messages
modelDisplayLabel: 'Groq'
​
iconURL
string
Custom icon URL for the endpoint
iconURL: https://example.com/icon.png

​
Headers Configuration

​
headers
object
Custom headers to include with requests
headers:
  x-portkey-api-key: '${PORTKEY_API_KEY}'
  x-portkey-virtual-key: '${PORTKEY_OPENAI_VIRTUAL_KEY}'
Special variables:
  • {{LIBRECHAT_BODY_PARENTMESSAGEID}} - Insert parent message ID

​
Request Parameters

​
addParams
object
Add or override default request parameters
addParams:
  safe_prompt: true
  temperature: 0.7
​
dropParams
array
Remove default parameters from requests
dropParams: ['stop', 'user', 'frequency_penalty', 'presence_penalty']
Required for some providers like Mistral to avoid 422 errors

​
Example: Groq

endpoints:
  custom:
    - name: 'groq'
      apiKey: '${GROQ_API_KEY}'
      baseURL: 'https://api.groq.com/openai/v1/'
      models:
        default:
          - 'llama3-70b-8192'
          - 'llama3-8b-8192'
          - 'mixtral-8x7b-32768'
          - 'gemma-7b-it'
        fetch: false
      titleConvo: true
      titleModel: 'mixtral-8x7b-32768'
      modelDisplayLabel: 'groq'

​
Example: Mistral AI

endpoints:
  custom:
    - name: 'Mistral'
      apiKey: '${MISTRAL_API_KEY}'
      baseURL: 'https://api.mistral.ai/v1'
      models:
        default: ['mistral-tiny', 'mistral-small', 'mistral-medium']
        fetch: true
      titleConvo: true
      titleModel: 'mistral-tiny'
      modelDisplayLabel: 'Mistral'
      dropParams: ['stop', 'user', 'frequency_penalty', 'presence_penalty']
Mistral AI’s API doesn’t support some OpenAI-standard parameters. Dropping them prevents 422 validation errors.

​
Example: OpenRouter

endpoints:
  custom:
    - name: 'OpenRouter'
      apiKey: '${OPENROUTER_KEY}'
      baseURL: 'https://openrouter.ai/api/v1'
      headers:
        x-librechat-body-parentmessageid: '{{LIBRECHAT_BODY_PARENTMESSAGEID}}'
      models:
        default: ['meta-llama/llama-3-70b-instruct']
        fetch: true
      titleConvo: true
      titleModel: 'meta-llama/llama-3-70b-instruct'
      dropParams: ['stop']
      modelDisplayLabel: 'OpenRouter'

​
Example: Helicone (AI Gateway)

endpoints:
  custom:
    - name: 'Helicone'
      apiKey: '${HELICONE_KEY}'
      baseURL: 'https://ai-gateway.helicone.ai'
      headers:
        x-librechat-body-parentmessageid: '{{LIBRECHAT_BODY_PARENTMESSAGEID}}'
      models:
        default:
          - 'gpt-4o-mini'
          - 'claude-4.5-sonnet'
          - 'llama-3.1-8b-instruct'
          - 'gemini-2.5-flash-lite'
        fetch: true
      titleConvo: true
      titleModel: 'gpt-4o-mini'
      modelDisplayLabel: 'Helicone'
      iconURL: https://marketing-assets-helicone.s3.us-west-2.amazonaws.com/helicone.png

​
Example: Portkey AI

endpoints:
  custom:
    - name: 'Portkey'
      apiKey: 'dummy'
      baseURL: 'https://api.portkey.ai/v1'
      headers:
        x-portkey-api-key: '${PORTKEY_API_KEY}'
        x-portkey-virtual-key: '${PORTKEY_OPENAI_VIRTUAL_KEY}'
      models:
        default: ['gpt-4o-mini', 'gpt-4o', 'chatgpt-4o-latest']
        fetch: true
      titleConvo: true
      titleModel: 'current_model'
      summarize: false
      summaryModel: 'current_model'
      modelDisplayLabel: 'Portkey'
      iconURL: https://images.crunchbase.com/image/upload/c_pad,f_auto,q_auto:eco,dpr_1/rjqy7ghvjoiu4cd1xjbf
current_model uses the currently selected model for titles/summaries

​
Azure OpenAI

While Azure OpenAI has a dedicated endpoint, you can also configure it as a custom endpoint: 
endpoints:
  azureOpenAI:
    # Azure-specific configuration here
Refer to the official Azure OpenAI documentation for detailed setup.

​
Model Specifications

Create preset configurations that can be grouped in the UI: 
modelSpecs:
  list:
    # Nested under endpoint
    - name: "gpt-4o"
      label: "GPT-4 Optimized"
      description: "Most capable GPT-4 model"
      group: "openAI"
      preset:
        endpoint: "openAI"
        model: "gpt-4o"
    
    # Custom group with icon
    - name: "coding-assistant"
      label: "Coding Assistant"
      description: "Specialized for coding tasks"
      group: "my-assistants"
      groupIcon: "https://example.com/icons/assistants.png"
      preset:
        endpoint: "openAI"
        model: "gpt-4o"
        instructions: "You are an expert coding assistant..."
        temperature: 0.3
    
    # Standalone (no group)
    - name: "general-assistant"
      label: "General Assistant"
      description: "General purpose assistant"
      preset:
        endpoint: "openAI"
        model: "gpt-4o-mini"
​
group
string
Organize specs in UI:
  • Match endpoint name to nest under that endpoint
  • Custom name creates separate collapsible section
  • Omit for standalone top-level item
​
groupIcon
string
Icon for custom groups (URL or built-in endpoint key)
Only needs to be set on one spec per group

​
Environment Variables

Set the following environment variables for your custom endpoints: 
# Groq
GROQ_API_KEY=your_groq_key

# Mistral AI
MISTRAL_API_KEY=your_mistral_key

# OpenRouter
OPENROUTER_KEY=your_openrouter_key

# Other providers
ANYSCALE_API_KEY=
COHERE_API_KEY=
DEEPSEEK_API_KEY=
FIREWORKS_API_KEY=
HUGGINGFACE_TOKEN=
PERPLEXITY_API_KEY=
TOGETHERAI_API_KEY=

​
Troubleshooting

Some providers don’t support all OpenAI parameters. Use dropParams to remove unsupported fields:
dropParams: ['stop', 'user', 'frequency_penalty', 'presence_penalty']
  1. Check models.default has at least one model
  2. Verify models.fetch: true if querying API
  3. Ensure API key is valid and has access
Make sure headers are under the headers key and use proper environment variable syntax:
headers:
  x-api-key: '${MY_API_KEY}'

Build docs developers (and LLMs) love

Get started for freeTalk to us