Gemini Canvas can turn a photo of a hand-drawn app idea into generated code for a prototype you can preview and refine. In a post published November 5, 2026, Google described using the sketch as a starting point, then asking Gemini to make changes in Canvas. It gives you a visual place to explore an app idea before building it out in code.
From sketch photo to Canvas prototype
Google’s examples include an app mock-up, a habit tracker, and a game board. The workflow is straightforward:
- Upload a photo of the sketch to Gemini and select Canvas.
- Ask Gemini, “Can you code this app?”
- Try the generated result in Preview.
- Ask for changes, such as different colors, fonts, or features.
Gemini interprets the drawing and generates code; Preview lets you try the resulting prototype, then refine it through further prompts.
Google Workspace had already demonstrated sketch upload, code generation, a live preview, and conversational edits in a video published October 27, 2025. The clip is in English.
Use the first generated version to explore the idea: the workflow is built around previewing it and asking for refinements.
How Workspace data affects preview
Google Workspace said in a May 20, 2025 update that an app containing Workspace data cannot be previewed. An app without Workspace data can be previewed, so this restriction applies to the data an app includes, not to every Canvas prototype.
Sharing Canvas work from Workspace accounts
The same Google Workspace update said business and education users cannot share Canvas-created content from their Workspace accounts. They can open a link shared by someone using a personal account, but they cannot then ask Gemini to refine that content.
The November 2026 post describes how to make a sketch-based prototype; Google Workspace’s October 2025 video had already shown the core workflow publicly.
