LibreChat integration

Add Modellane to LibreChat as a custom endpoint in librechat.yaml, with the API key read from your environment file.

LibreChat supports custom OpenAI-compatible endpoints. Declare Modellane in librechat.yaml and its models appear in the endpoint menu.

LibreChat is a self-hosted chat interface. It supports custom OpenAI-compatible endpoints, which call the Chat Completions API (POST /v1/chat/completions).

Add the endpoint#

Put the key in LibreChat's .env file:

Shell

MODELLANE_API_KEY=your-api-key

Then add Modellane under endpoints.custom in librechat.yaml:

YAML

endpoints:
  custom:
    - name: "Modellane"
      apiKey: "${MODELLANE_API_KEY}"
      baseURL: "https://usemodellane.com/v1"
      models:
        default: ["lane-1"]
        fetch: true
      titleConvo: true
      titleModel: "lane-1"
      modelDisplayLabel: "Modellane"
ParamValue
apiKeyThe variable in .env that holds your key from API keys
baseURLhttps://usemodellane.com/v1
models.defaultThe models shown in the menu, ids from Models & pricing
models.fetchtrue loads every model id from GET /v1/models
titleModelThe model that writes conversation titles; a small, cheap model is enough

Restart LibreChat so it reads the new configuration, then pick Modellane in the endpoint menu.

If a model rejects one of the sampling parameters LibreChat sends, list it under dropParams for the endpoint, for example dropParams: ["frequency_penalty", "presence_penalty"].

Verify#

Start a new conversation with the Modellane endpoint and send "Reply with the word ready." A reply means the endpoint works. An authentication error points at the key, and model_not_found at the model id.

Images#