SketchMind unveils SketchGen 2.0, a multimodal UI generator for rapid prototypes

AI · 4 min read

SketchGen 2.0 adds multimodal inputs—hand-drawn sketches, voice prompts, and screenshots—to its generative UI engine, enabling designers to move from rough concepts to workable prototypes in minutes. The model understands design tokens and can align generated components with an existing design system to reduce style drift.

The update introduces 'intent refinement', an interactive loop where designers accept, edit, or request alternatives and the model re-generates contextually consistent variations. Exports include editable Figma files, React components with storybook stubs, and accessibility annotations that flag contrast and semantic issues.

SketchMind also introduced enterprise controls for IP protection and on-prem model options aimed at regulated industries. The company plans additional integrations with usability testing platforms so generated prototypes can be validated with users without manual conversion.