1 min read
Add as a preferred source on Google

Google Adds Sketch-Based App Building to Gemini’s Canvas Tool

Users can turn a photo of a hand-drawn interface into code, then test and revise the app in Canvas.

A pencil app sketch rests beside a tablet interface preview / TokenPost.ai
A pencil app sketch rests beside a tablet interface preview / TokenPost.ai

Google added sketch input to Gemini’s Canvas tool on Nov. 5, letting users turn photographed app drawings into editable code and interactive previews for prototype creation.

Users can upload a picture of an interface and ask Gemini to build it. Canvas generates code and displays a preview where users can try the result and adjust details such as colors, fonts and features.

Canvas automatically provisions a Firestore database instance when an app needs persistent data, though users can select another storage option. People with a project’s public link can view and edit its shared data.

Canvas debuted in March 2025 and gained Gemini 2.5 models in May 2025. It already generated HTML and React web prototypes, along with Python scripts, from text prompts. Sketches now provide another way to start those projects.

Simon Yoon

Reporter

Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to info@tokenpost.com.

Loading…