Help Center

LLM Gateway

One OpenAI-compatible endpoint and one key for every model, with your usage in one place.

The gateway gives you a single OpenAI-compatible base URL and key that reaches many models across providers — so switching model is a string change, not an integration.

Because every call goes through one place, spend and usage are visible per key rather than scattered across provider dashboards.

How to

  1. 1

    Create a gateway key.

  2. 2

    Point any OpenAI-compatible client at the gateway base URL with that key.

  3. 3

    Pick a model by name.

  4. 4

    Watch usage per key.

If something goes wrong

A model name is rejected.
Not every provider model is exposed. Check the model list in the dashboard rather than assuming a provider name works.

Related