GetRouterDocs
  • English

    • OpenAI-compatible Tools

    Use this guide when a tool offers OpenAI-compatible, Custom OpenAI, or Custom Provider settings and sends Chat Completions requests.

    Integration status
    • Recommended protocol: OpenAI Chat Completions
    • Configuration checked: July 15, 2026
    • Runtime verification: Depends on the client and version

    Configuration

    Provider name: GetRouter
API key:       YOUR_API_KEY
Base URL:      https://getrouter.ai/v1
Model:         MODEL_ID

    Use an exact model ID from Models.

    URL behavior

    Tools do not agree on what “Base URL” means:

    • If the tool appends /chat/completions, enter https://getrouter.ai/v1.
    • If the tool appends /v1/chat/completions, it may expect https://getrouter.ai.
    • If the field asks for a complete endpoint, enter https://getrouter.ai/v1/chat/completions.

    Do not guess. Check the field description or the tool's network/error output.

    Verify

    1. Save the provider.
    2. Select MODEL_ID explicitly if the model list is empty.
    3. Send a short text message.
    4. Confirm the request appears in GetRouter usage logs.
    5. Test streaming and tool calling separately if you need them.

    Common errors

    SymptomCheck
    401API key, spaces, expired or disabled key
    404Repeated /v1, wrong operation path, wrong protocol
    Model not foundExact model ID and key permissions
    Models load but chat failsThe model may not support Chat Completions
    Chat works but tools failModel or client may not support compatible tool calls

    See Connection troubleshooting for a complete checklist.