One API for your chat completions
Puma is an OpenRouter-compatible chat completions API powered by Cloudflare Workers AI. Drop it into any OpenAI client — just change the base URL.
Drop-in compatible
Speaks the OpenAI chat completions protocol — point any client at Puma by changing the base URL.
Edge-fast
Runs on Cloudflare Workers AI, so requests are handled close to your users, anywhere in the world.
One endpoint
A single /api/v1/chat/completions route. No SDK, no boilerplate — just HTTP and a Bearer token.
Quick start
curl https://api.pumaai.com/api/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer sk-test-123" \
-d '{"messages":[{"role":"user","content":"Hello!"}]}'