Models and pricing
Choose a model and check its current price, context window, and supported features.
Supported models
Loading models
| Model | Input | Cached input | Output | Context | Copy model ID |
|---|
If you want to load the available models on demand, you can call GET https://api.gradiated.com/api/v1/models.
This endpoint also returns output limits, streaming support, and supported parameters.
Each model's properties.type identifies its inference interface: chat for text generation or decision for typed question scoring. Filter on properties.type === "chat" when building a Chat Completions model list.
Each model's properties.supported_uses lists what it can be used for. Only models that include serving accept inference requests; the catalog also lists training bases that you can fine-tune but not call directly.