GET /v1/models to list available models and GET /v1/models/{id} to inspect one model. Discovery does not generate tokens or charge credits.
All models support Chat Completions, Responses, Messages, streaming, function tools, structured outputs, and policies. Image inputs are supported only on models that advertise them. Video is not supported. See compatibility.
Input and output share the context window. Large output limits do not guarantee a response of that length; timeouts and remaining context still apply. Use the live catalog for current limits and reasoning settings.