Getting started

Three short pages take you from nothing to your first picture:

1. Install and check your phone — getting the app, and finding out which models your chip can run.
2. Your first picture — download one model, open a ready-made flow, press Run.
3. The canvas — what everything on the main screen does: nodes, wires, gestures, the run log.

If you have used ComfyUI before, the idea is the same: nodes joined by wires. The difference is that each node here does a whole job (a "Text to image" node encodes the prompt, samples and decodes by itself), so a complete flow is usually three or four nodes, not fifteen.