POST /v1/chat/completions

OpenAI-compatible chat. Accepts the standard request schema, returns the standard response shape, adds `model` field showing which one it routed to.

POST/api/v1/chat/completions

Standard OpenAI-compatible completions with smart routing.

Parameters

Name	Type	Required	Description
`messages`	`array`	yes	OpenAI-format message list
`model`	`string`	no	`auto` to let Soriku route, or a specific model id
`stream`	`boolean`	no	Stream tokens via SSE
`temperature`	`number`	no	Sampling temperature
`max_tokens`	`integer`	no	Cap on response length
`x_agent_id`	`string`	no	Attach an agent persona for this request

Example

curl -X POST http://localhost:8765/api/v1/chat/completions \
  -H "Authorization: Bearer soriku-local" \
  -H "Content-Type: application/json" \
  -d '{"model":"auto","messages":[{"role":"user","content":"Hello"}]}'

Response

{
  "id": "chatcmpl-...",
  "object": "chat.completion",
  "model": "qwen2.5-coder:7b",
  "choices": [{
    "message": {"role": "assistant", "content": "..."},
    "finish_reason": "stop"
  }]
}