Screenshot to Code
AI web app that turns screenshots and mockups into front-end code
What it is
A web app with a React frontend and FastAPI backend that sends screenshots, Figma designs, mockups, or screen recordings to vision models from OpenAI, Anthropic, or Google and returns working front-end code. Output can be HTML with Tailwind or CSS, React, Vue, Bootstrap, or Ionic, and the agent can render its result in a headless browser to check it against the source. Front-end developers and designers run it to get a quick starting prototype of an existing page or design instead of hand-coding the layout. It runs locally or via Docker with your own API keys, or as a hosted service.