LibreChat
Features

Query Parameters

Learn how to configure chat conversations using URL query parameters in LibreChat. Set models, endpoints, and conversation settings dynamically.

LibreChat supports dynamic configuration of chat conversations through URL query parameters. This feature allows you to initiate conversations with specific settings, models, and endpoints directly from the URL.

Chat Paths

Query parameters must follow a valid chat path:

  • For new conversations: /c/new?
  • For existing conversations: /c/[conversation-id]? (where conversation-id is an existing one)

Examples:

https://your-domain.com/c/new?endpoint=ollama&model=llama3%3Alatest
https://your-domain.com/c/03debefd-6a50-438a-904d-1a806f82aad4?endpoint=openAI&model=o1-mini

Basic Usage

The most common parameters to use are endpoint and model. Using both is recommended for the most predictable behavior:

https://your-domain.com/c/new?endpoint=azureOpenAI&model=o1-mini

URL Encoding

Special characters in query params must be properly URL-encoded to work correctly. Common characters that need encoding:

  • : → %3A
  • / → %2F
  • ? → %3F
  • # → %23
  • & → %26
  • = → %3D
  • + → %2B
  • Space → %20 (or +)

Example with special characters:

Original: `Write a function: def hello()`
Encoded: `/c/new?prompt=Write%20a%20function%3A%20def%20hello()`

You can use JavaScript's built-in encodeURIComponent() function to properly encode prompts:

const prompt = "Write a function: def hello()";
const encodedPrompt = encodeURIComponent(prompt);
const url = `/c/new?prompt=${encodedPrompt}`;
console.log(url);

Try running the code in your browser console to see the encoded URL (browser shortcut: Ctrl+Shift+I).

Endpoint Selection

The endpoint parameter can be used alone:

https://your-domain.com/c/new?endpoint=google

When only endpoint is specified:

  • It will use the last selected model from localStorage
  • If no previous model exists, it will use the first available model in the endpoint's model list

Notes

  • The endpoint value must be one of the following:
openAI, azureOpenAI, google, anthropic, assistants, azureAssistants, bedrock, agents
  • If using a custom endpoint, you can use its name as the value (case-insensitive)
# using `endpoint=perplexity` for a custom endpoint named `Perplexity`
https://your-domain.com/c/new?endpoint=perplexity&model=llama-3.1-sonar-small-128k-online

Model Selection

The model parameter can be used alone:

https://your-domain.com/c/new?model=gpt-4o

When only model is specified:

  • It will only select the model if it's available in the current endpoint
  • The current endpoint is either the default endpoint or the last selected endpoint

Prompt Parameter

The prompt parameter allows you to pre-populate the chat input field:

https://your-domain.com/c/new?prompt=Explain quantum computing

You can also use q as a shorthand, which is interchangeable with prompt:

https://your-domain.com/c/new?q=Explain quantum computing

You can combine these with other parameters:

https://your-domain.com/c/new?endpoint=anthropic&model=claude-3-5-sonnet-20241022&prompt=Explain quantum computing

Automatic Prompt Submission

The submit parameter allows you to automatically submit the prompt without manual intervention:

https://your-domain.com/c/new?prompt=Explain quantum computing&submit=true

This feature is particularly useful for:

  • Creating automated workflows (e.g., Raycast, Alfred, Automater)
  • Building external integrations

You can combine it with other parameters for complete automation:

https://your-domain.com/c/new?endpoint=openAI&model=gpt-4&prompt=Explain quantum computing&submit=true

Special Endpoints

Model Specs

You can select a specific model spec by name:

https://your-domain.com/c/new?spec=meeting-notes-gpt4

This will load all the settings defined in the model spec. When using the spec parameter, other model parameters in the URL will be ignored.

Agents

You can directly load an agent using its ID without specifying the endpoint:

https://your-domain.com/c/new?agent_id=your-agent-id

This will automatically set the endpoint to agents.

Assistants

Similarly, you can load an assistant directly:

https://your-domain.com/c/new?assistant_id=your-assistant-id

This will automatically set the endpoint to assistants.

Supported Parameters

LibreChat supports a wide range of parameters for fine-tuning your conversation settings:

LibreChat Settings

  • maxContextTokens: Override the system-defined context window
  • resendFiles: Control file resubmission in subsequent messages
  • promptPrefix: Set custom instructions/system message
  • imageDetail: 'low', 'auto', or 'high' for image quality
    • Note: while this is a LibreChat-specific parameter, it only affects the following endpoints:
    • OpenAI, Custom Endpoints, which are OpenAI-like, and Azure OpenAI, for which this defaults to 'auto'
  • spec: Select a specific LibreChat Model Spec by name
    • Must match the exact name of a configured model spec
    • When specified, other model parameters will not take effect, only those defined by the model spec
    • Important: If model specs are configured with enforce: true, using this parameter may be required for URL query params to work properly
  • fileTokenLimit: Set maximum token limit for file processing to control costs and resource usage.
    • Note: Request value overrides YAML default.

Model Parameters

Different endpoints support various parameters:

OpenAI, Custom, Azure OpenAI:

# Note: these should be valid values according to the provider's API
temperature, presence_penalty, frequency_penalty, stop, top_p, max_tokens,
reasoning_effort, reasoning_summary, verbosity, useResponsesApi, web_search, disableStreaming

Google, Anthropic:

# Note: these should be valid values according to the provider's API
topP, topK, maxOutputTokens, thinking, thinkingBudget, web_search

Anthropic, Bedrock (Anthropic models):

Set this to true or false to toggle the "prompt-caching":

promptCache

More info: https://www.anthropic.com/news/prompt-caching, https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html#prompt-caching-get-started

Bedrock:

# Bedrock region
region=us-west-2
# Bedrock equivalent of `max_tokens`
maxTokens=200

Assistants/Azure Assistants:

# overrides existing assistant instructions for current run
instructions=your+instructions
# Adds the current date and time to `additional_instructions` for each run.
append_current_datetime=true

More Info

For more information on any of the above, refer to Model Spec Preset Fields, which shares most parameters.

Example with multiple parameters:

https://your-domain.com/c/new?endpoint=google&model=gemini-2.0-flash-exp&temperature=0.7&prompt=Oh hi mark

Example with model spec:

https://your-domain.com/c/new?spec=meeting-notes-gpt4&prompt=Here%20is%20the%20transcript...

Note: When using spec, other model parameters are ignored in favor of the model spec's configuration.

⚠️ Warning

Exercise caution when using query parameters:

  • Misuse or exceeding provider limits may result in API errors
  • If you encounter bad request errors, reset the conversation by clicking "New Chat"
  • Some parameters may have no effect if they're not supported by the selected endpoint

Best Practices

  1. Always use both endpoint and model when possible
  2. Verify parameter support for your chosen endpoint
  3. Use reasonable values within provider limits
  4. Test your parameter combinations before sharing URLs

Parameter Validation

All parameters are validated against LibreChat's schema before being applied. Invalid parameters or values will be ignored, and valid settings will be applied to the conversation.

This feature enables powerful use cases like:

  • Sharing specific conversation configurations
  • Creating bookmarks for different chat settings
  • Automating chat setup through URL parameters

#LibreChat #ChatConfiguration #AIParameters #OpenSource #URLQueryParameters

How is this guide?

Edit on GitHub

Previous

Temporary Chat

Next

Import Conversations

On this page