Model API

Kanonas

Outsource your evals. Send one API request and let Kanonas learn which model should handle each task.

Accounts
Docs
API keys
Trace views

Account console

Bring your keys. Call one gateway.

Sign in with Google, save provider credentials, create a Kanonas API key, then call OpenAI-, Anthropic-, or Gemini-shaped endpoints.

1Google account
2Provider keys
3Kanonas API key
curl https://api.kanonas.ai/v1/chat/completions \
  -H "Authorization: Bearer key_..." \
  -H "Content-Type: application/json" \
  -d '{"model":"anthropic/claude-3-5-haiku-20241022","messages":[{"role":"user","content":"Say ack"}],"max_tokens":32}'
curl https://api.kanonas.ai/v1/messages \
  -H "Authorization: Bearer key_..." \
  -H "Content-Type: application/json" \
  -d '{"model":"openai/gpt-4o-mini","max_tokens":32,"messages":[{"role":"user","content":"Say ack"}]}'
curl https://api.kanonas.ai/v1beta/models/openai/gpt-4o-mini:generateContent \
  -H "Authorization: Bearer key_..." \
  -H "Content-Type: application/json" \
  -d '{"contents":[{"role":"user","parts":[{"text":"Say ack"}]}],"generationConfig":{"maxOutputTokens":32}}'

OpenAI-compatible endpoints first. Routing, eval replay, and traces next.