Inference API/Text Generation
Create a response
Creates a model response with the stateless Responses API (OpenAI-compatible); send the full conversation in input on every request, with optional server-sent-event streaming.
Create a chat completion POST
Creates a model response for a chat conversation (OpenAI-compatible), with optional server-sent-event streaming, a final usage chunk by default, and reasoning text returned in reasoning_content.
Create a message (Anthropic-compatible) POST
Creates a message with the Anthropic-compatible Messages API, for the Anthropic SDKs and tools built on them, with optional streaming.