Batch Image Generation With Queue Management And Resource Pooling

Professional open-source creative engine with node-based workflow editor.

Unique: Implements an in-memory invocation queue with priority support and automatic resource pooling that unloads unused models to maximize GPU utilization. Queue status is exposed via REST API with real-time updates via WebSocket events.

vs others: Simpler than external job queue systems (Celery, RQ) because it's built into the FastAPI application, while more efficient than naive sequential processing because it can batch similar generations and manage model loading intelligently.