ModelGateway documentation
One API. Every route.
Integrate once to discover models, route inference across providers, and manage marketplace supply.
Make your first request
Create a key, discover available models, and run an OpenAI-compatible chat request.
Understand the router
See how eligibility, health, capacity, marketplace price, and fallback determine a route.
Choose a model
Discover the live catalog and understand model IDs and endpoint compatibility.
Integrate an API
Use OpenAI-compatible endpoints or keep native Anthropic and Gemini request formats.