OpenStitch, open-source AI UI prototyping tool that runs locally with Ollama
https://reddit.com/link/1sci25b/video/fpqaqqnjn6tg1/player
Built this over the past few days. You describe a screen (or drop a screenshot, or sketch a wireframe) and it generates rendered, interactive frontend code on an infinite canvas. Link screens into flows and prototype them in-app.
Runs fully local with Ollama. No cloud, no accounts. OpenRouter works too if you want stronger vision models.
Main workflows:
- Generate: describe a full product, get multiple screens with a shared design system
- Screenshot to UI: drop a screenshot or wireframe sketch, get a code replica
- Iterate: refine any screen with follow-up prompts
Stack: React + FastAPI + SQLite + Ollama. Runs via Docker Compose.
Tested with Qwen3-coder:30b for code and Qwen3.5-122B-A10B for vision.