Stability releases StableUI: a diffusion-powered layout assistant for screen design

AI · 4 min read

Stability AI extended its generative portfolio with StableUI, a model that uses diffusion techniques to synthesize entire screen designs and adaptive layouts. Users can provide a text brief, a rough wireframe sketch, or brand tokens to obtain multiple candidate interfaces.

Unlike image-only generators, StableUI produces structured outputs: layers, text blocks, and suggested components that are exportable to Figma files or as HTML/CSS snippets. The model includes presets for platforms like mobile, tablet, and desktop, and supports accessibility annotations for contrast and focus order.

Early adopters praised the model for quickly exploring visual directions, though they cautioned that real product work still requires human refinement. Stability plans a plugin ecosystem for additional integrations and is shipping an enterprise tier with model fine-tuning on private brand assets.