Skip to main content

OpenAI Chat Completions-compatible text generation

POST 
/v1/chat/completions

Generate text with an OpenAI Chat Completions-compatible payload. The requested model must already be placed, running, and declare TextGeneration; speech-only models are rejected before command dispatch. The reserved model id 'skulk/steward' selects the intelligent-fabric steward and answers 404 when that mode is disabled, or 503 with the steward status payload while the steward is still being placed, staged, or loaded.

Request​

Responses​

Successful Response