Moon — Routing · fastest healthy model

Route through the fastest healthy model, every request.

Moon watches live p95 latency and error rate for every model in your chain. When the leader slows down, traffic shifts in under one second.

How it works.

Three moving parts, one clean API call.

01
Declare your chain

Pass a primary model and up to four fallbacks. Moon records both live latency and error rate per model.

02
Watch and decide

Every request checks the current p95 against your budget. If the leader is slow, Moon picks the next healthy option instantly.

03
See which one served

Every trace records which model actually served the request under a served_by field. The fallback rate shows up in your overview.

Why teams use it.

Reasons users tell us they turn on routing on day one.

Cuts p95 latency in half.When the primary model has a bad minute, traffic finishes on a warm replica.
Zero markup on routing.Fallback chains cost the same as a single-model call. Only pay for what serves.
Region-aware.Route iad first, lhr on failover, sfo on failure. All from one config line.
Works with any provider.OpenAI, Anthropic, Google, and hosted open weights. Bring your own key.

Try it now.

One config line. Every SDK.

javascriptCopy
await moon.chat.stream({
  model: "claude-sonnet-5",
  fallback: ["claude-haiku-4-5", "gpt-4.2-mini"],
  budget_ms: 800,
});
// Every trace records which model actually served
© 2026 Moon · Privacy · Terms · Status
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon
© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon© BrowserMoon