Skip to content

LLM gateway

What is an LLM gateway?

An LLM gateway is a proxy between your app and many model providers. You call one base URL with one API key, and the gateway routes the request, retries it on another provider when the first fails, and logs it in one place.

What an AI router does for you

One API

OpenAI Chat Completions, Responses, Embeddings and Anthropic Messages behind one base URL.

Automatic failover

A 401, 429 or 5xx from one upstream retries on the next, so the caller sees the answer, not the error.

Free models

A catalog of $0-per-token models on Go and up, with no daily cap and no credit deduction.

Your own keys, $0 markup

Route through provider keys you already pay for and the markup is $0.

One log for every call

Every request records model, tokens, cost, latency and errors in one place.

Honest comparisons

Fees, free-model limits and plans compared with sources and dates.

How routing works

  1. 1.Send a request to the gateway base URL with any model id, written provider/model.
  2. 2.The gateway picks an upstream for that model, from your own keys first when you have them.
  3. 3.On a 401, 429 or 5xx it retries the next upstream, and your client sees the final answer.