POST /api/v1/playground/stream

Stream an LLM completion using a project's configured provider key. Used internally by the dashboard playground, but available to any authenticated client.

Request

POST /api/v1/playground/stream
Authorization: Bearer ts_...
Content-Type: application/json

{
  "model": "gpt-4o",
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user", "content": "Explain quicksort in one paragraph." }
  ],
  "temperature": 0.7,
  "max_tokens": 500
}

Parameters

Response

Returns a Server-Sent Events (SSE) stream. Each event contains a chunk of the completion:

event: token
data: {"content": "Quick"}

event: token
data: {"content": "sort"}

event: done
data: {"usage": {"prompt_tokens": 25, "completion_tokens": 150, "total_tokens": 175}}

Requirements

• The project must have an LLM provider key configured for the requested model's provider.
• Provider keys are managed via the llmKey tRPC router or the Project Settings page in the dashboard.

Errors