Skip to content
AnyRouter

Every LLM.
One API.

OpenAI- and Anthropic-compatible. Switch models, keep your code.

QuickstartBrowse 258 models
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "anyrouter/auto", "messages": [{"role": "user", "content": "Hi"}]}'
Hi! I'm here and ready to help. What can I do for you today? Ask me anything, from code and writing to quick facts.
ExpandModelContextPrice
Nemotron 3.5 Super VL 120B A12BFree1MPool
Muse Wonder1MBYOK$4.00 in · $20.00 out
Amos 1 LargeFree131KUpstream
Amos 1 SmallFree262KUpstream
THX-01Free1KUpstream
Microsoft-Decision-1Free33KPool
Drex v1.5Free131KPool
Solar Decide FlashFree524KPool
Browse all 258 models

How a request routes

Your app sends one request to AnyRouter. AnyRouter checks the key, picks a route and, if an upstream fails, falls back to the next one. You are billed once.

Fig. 1POST /v1/chat/completions
CLI
IDE
Your app
AnyRouterkey · route · fallback
Workers AI429 → fallback
OpenRouter200
CLI
IDE
Your app
AnyRouterkey · route · fallback
Workers AI429 → fallback
OpenRouter200
  1. 1Apps send POST /v1/chat/completions
  2. 2AnyRouter checks the key, picks a route by model
  3. 3Upstream 1 fails, falls back to upstream 2
  4. 4The answer streams back
429 → fallback · 200 · billed once

FAQ

What is AnyRouter?

OpenAI-compatible gateway. 258 models, 17 providers. Free models, $4 credit on Go, $0 markup on your own keys.

How do I call models through AnyRouter?

Point your SDK at one base URL and reference models as provider/model, as in the code above. The table lists every model with its price per 1 million tokens.

Do you take a cut if I bring my own keys?

No. Route through your own provider keys and the markup is $0: what you pay upstream is exactly what you pay.

How do I get started?

Sign up in seconds, no card required. Free models cost nothing, Go adds $4 credit, and the table above shows what every other model costs.