Skip to main content
The Miteos API gives you programmatic access to the full multi-agent orchestration engine. Submit natural language tasks, control running agents, stream real-time events over WebSocket, and integrate AI-driven workflows directly into your product — all through a single REST interface.

Base URL

Every API request is made to this base URL:
https://api.miteos.com/v1
All endpoints are versioned under /v1. Include this prefix in every request path.

What You Can Do

Submit Tasks

Send a natural language prompt and let the orchestrator automatically decompose it into a multi-agent execution plan.

Control Agents

Pause, resume, guide, or kill agents mid-execution for full real-time control over your agent fleet.

Real-time Events

Stream every agent action, approval request, and deliverable update over WebSocket or receive them as webhook notifications.

Deploy Outputs

Deploy agent-built websites, apps, and services to Vercel, Cloudflare, Expo, or the App Store with a single API call.

SDKs

Official SDKs handle authentication, retries, streaming, and type safety out of the box. Install the one that matches your stack:
npm install @miteos/sdk

Rate Limits

Rate limits are enforced per API key. If you exceed your plan’s limit, the API returns a 429 Too Many Requests response. The X-RateLimit-Limit and X-RateLimit-Remaining headers on every response show your current usage.
TierRequests / minConcurrent TasksWebSocket Connections
Free6011
Pro30055
Power1,0002020
Business5,000Unlimited100
EnterpriseCustomCustomCustom

Error Responses

All errors return a JSON body with a single detail field containing a human-readable message:
{
  "detail": "Human-readable error message"
}
Use the HTTP status code to determine how to handle the error programmatically:
StatusNameMeaning
400Bad RequestThe request body or parameters are malformed.
401UnauthorizedMissing or invalid API key.
403ForbiddenYour key does not have the required scope for this action.
404Not FoundThe requested resource does not exist.
422Validation ErrorThe request is well-formed but fails business logic validation.
429Rate LimitedYou have exceeded your plan’s request rate limit.
500Server ErrorAn unexpected error occurred on Miteos servers.

OpenAPI Spec

The full machine-readable OpenAPI specification is available at:
https://api.miteos.com/v1/openapi.json
Use it to generate client code, import into Postman or Insomnia, or power your own documentation tooling.

Quick Example

The following curl request creates a new task and returns the task object, including its id which you can use to poll for status or stream events:
curl -X POST https://api.miteos.com/v1/tasks \
  -H "Authorization: Bearer mt_live_xxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Research our top 5 competitors and build a pricing comparison landing page",
    "config": {
      "max_agents": 3,
      "autonomy_level": 80,
      "auto_deploy": true,
      "deploy_target": "vercel"
    }
  }'
A successful response looks like this:
{
  "id": "task_x9y8z7w6v5",
  "status": "planning",
  "prompt": "Research our top 5 competitors and build a pricing comparison landing page",
  "created_at": "2026-06-18T10:44:00Z",
  "config": {
    "max_agents": 3,
    "autonomy_level": 80,
    "auto_deploy": true,
    "deploy_target": "vercel"
  }
}
Use the returned id to stream real-time events or retrieve deliverables once the task completes.