Key Features

Recovers editable design structure from static images.
Uses agentic decomposition to identify components and layout hierarchy.
Targets UI, graphic design, and Figma-style reconstruction workflows.
Provides an official public GitHub repository.
Links to a Figma-focused dataset on Hugging Face.
Includes a teaser video demonstrating the recovered design workflow.
Supports research into design automation and image-to-layout conversion.
Useful for making screenshot-based designs easier to inspect and edit.

The system decomposes visual layouts into meaningful design components, using agentic reasoning to identify text, shapes, hierarchy, and layout relationships. Its goal is to bridge static images and editable design representations for workflows such as Figma reconstruction, UI analysis, and design asset recovery.


ReDesign is useful for design automation, UI reverse engineering, dataset creation, and creative tooling. Developers can use the repository to study how agentic decomposition can recover reusable design structure from images that would otherwise be difficult to edit directly.

Get more likes & reach the top of search results by adding this button on your site!

Embed button preview - Light theme
Embed button preview - Dark theme
TurboType Banner

Subscribe to the AI Search Newsletter

Get top updates in AI to your inbox every weekend. It's free!