Skip to content
Auteryn

API Reference

Agent API base URL: https://agent-api.auteryn.ai

Auth API base URL: https://profile-api.auteryn.ai

The Auteryn REST API lets you integrate AI agents directly into your applications.

Endpoints

Section titled “Endpoints”

POST /api/run

Run an agent — SSE stream (default) or dispatch to the background. Reference →

POST /api/run/async

Fire-and-forget background runs; poll, stream, or reconnect. Reference →

GET /api/tasks

List, get, control, and stream tasks and runs — incl. HITL input. Reference →

GET /api/tasks/{id}/messages

Retrieve message history for a task. Reference →

GET /api/agents/{id}/memories

Read agent long-term memory. Reference →

Errors & rate limits

Status codes, error shape, and 429 handling. Reference →

POST /api/v1/widget/*

Widget gateway for embedded chat. Reference →

Authentication

Section titled “Authentication”
  1. Create an org API key (sk_live_...) in the Console
  2. Exchange it for a JWT via POST /api/v1/auth/exchange/agent-token on the auth API
  3. Pass Authorization: Bearer <jwt> and X-Agent-Id on agent API requests

Full authentication guide →

Rate limits

Section titled “Rate limits”

Rate limits are enforced per org+user at the agent API layer (default 30/min for POST /api/run). On 429 Too Many Requests, honor the Retry-After header and back off. See the Errors & rate limits reference →.

Streaming

Section titled “Streaming”

POST /api/run returns Server-Sent Events (SSE) by default — real-time frames for agent text, tool calls, artifacts, and status. Each data: frame is one JSON event; see the AG-UI event types.

For background execution, set "stream": false (returns 202) or use POST /api/run/async, then follow progress with GET /api/tasks/{task_id}/stream (SSE) or GET /api/runs/{run_id}/events (paginated JSON).