Upstream APIs

InvokeAI server API reference.

API Endpoints

Images

• GET /api/v1/images/ - List images
• POST /api/v1/images/upload - Upload image
• GET /api/v1/images/i/{name} - Get image
• DELETE /api/v1/images/i/{name} - Delete image

Boards

• GET /api/v1/boards/ - List boards
• POST /api/v1/boards/ - Create board
• GET /api/v1/boards/{id} - Get board
• DELETE /api/v1/boards/{id} - Delete board

Models

• GET /api/v1/models/ - List models
• GET /api/v1/models/i/{key} - Get model info
• POST /api/v1/models/load - Load model

Queue

• POST /api/v1/queue/default/enqueue_batch - Submit batch
• GET /api/v1/queue/default/status - Queue status
• GET /api/v1/queue/default/session/{id} - Session status

WebSocket Events

// Connection
ws://localhost:9090/ws

// Events
{
  "event": "invocation_started",
  "data": {...}
}

{
  "event": "invocation_complete",
  "data": {...}
}

{
  "event": "session_complete",
  "data": {...}
}

OpenAPI Specification

Full API spec available at: - Development: http://localhost:9090/openapi.json - Reference: context/hints/invokeai-kb/invokeai-openapi.json

Rate Limiting

• Default: 100 requests/minute
• Batch operations: 10 requests/minute
• WebSocket: 1 connection per client

See API Reference for client-side APIs.