Model routingruntime
Give your agent 300+ models through one API
Name any model. Vetta routes it. Swap providers without changing the agent.
router300+
POST /v1/chat → one endpoint
OpenAI84Anthropic62Google48Meta36Mistral28
Routed to
claude-sonnet-4.6anthropic / chat completions
Streaming248 tokens · $0.004
Next in chainauto-fallback
- OpenAIgpt-5queued
- Googlegemini-2.5-proqueued
- Metallama-4queued
- DeepSeekdeepseek-v3queued
Paste and ship
Drop this into Claude Code, Codex, or Cursor. It reads the docs and wires up Model routing for you.
Agent prompt
Read https://vetta.sh/docs/concepts/model-router and use the naïve Model routing primitive in my project. It is part of the runtime layer; wire it through naive.config.ts rather than calling a vendor directly.