OpenRouter Model Gateway
A unified model API for applications that need provider routing, fallback handling and a choice of language models without separate integrations for each provider.
About this listing
OpenRouter provides a common API for calling models from multiple providers. Its routing layer can handle fallback attempts, making it useful for agent applications that need alternatives when a chosen provider cannot serve a request.
Developers can send HTTP requests directly or use the OpenAI SDK with OpenRouter's base URL. The documentation also describes streaming and a model-list endpoint, so applications can discover available model identifiers without hard-coding a catalogue.
Paid inference is metered at the selected model's published input and output token rates, using account credits. Credit purchases carry a platform fee. A limited free-model tier is available, but this does not make paid model inference free; model availability, limits and plan terms should be checked before deployment.