Models and Playground
Browse every model you can call with prices and capabilities, and try models side by side without writing code.
Both live under AI Gateway in a project's sidebar.
Models
AI Gateway → Models lists every model you can call through Cencori, with its provider, context window and price per million input and output tokens.
- Search by model name, or narrow the list to one or more providers.
- Filter by capability: Reasoning, Vision (image input), Code, Web search, Image generation, Tool use, Structured output, File input, Video input, Audio input and Prompt caching.
The model ID shown is what you pass as model in a request. The same catalog, with live availability, is on the Models page and in GET /v1/models.
Playground
AI Gateway → Playground is a chat for trying models before you write code.
- Pick a model from the composer, then ask a question.
- Compare models: add up to 8 models and each answers the same message side by side. Use it to weigh quality, speed and cost before you choose.
- Model configuration: temperature, max tokens, presence penalty and frequency penalty, per model.
- Regenerate an answer, copy it, or start a new chat.
Playground requests run through your project like any other request: they're billed to your credits and appear in Logs.