Install the local image model
This generator needs to install SD-Turbo (~2.3 GB) into your browser cache. The full diffusion model runs on your GPU β no backend, no API key, no third-party endpoint. After installing, generation takes a couple of seconds per image and works offline.
Stored in your browser's Cache Storage. You can clear it anytime from browser settings. First load takes a minute or two; everything after that runs from cache.
Type a prompt and hit Generate.
Each image is 512Γ512 and takes roughly 1β5 seconds on a discrete GPU.