Google has announced that users of the Gemini app can upload hand-drawn sketches to its Canvas tool and have Gemini generate functional app code from them, with the option to refine that code in the same workflow.
Google · Gemini app · Canvas
From paper sketch to working app code
Gemini users can upload a hand-drawn app sketch to Canvas, generate functional code from it, and refine that code in the same workflow. Google frames it as a faster path from rough idea to running prototype.
How the workflow runs
Step 1
Sketch
Upload a hand-drawn app layout to Canvas.
→
Step 2
Interpret
Gemini reads the image alongside text prompts to infer layout and behavior.
→
Step 3
Generate
Code is produced, not a static mockup, so it can be run directly.
→
Step 4
Refine
Adjust the app through further instructions in the same session.
Potential upside
Speed: code is generated and refined quickly once a sketch is uploaded.
Lower barrier: people with little or no programming background can build prototypes.
Shorter early experiments for designers, product managers and hobbyists.
Open questions
Accuracy on complex or ambiguous drawings is not detailed.
Supported languages and frameworks are not specified.
Limits on app complexity are not stated.
No independent evaluation of output quality was included.
Bottom line
Promising for prototyping; production reliability is still unproven.
Users will likely need to review and test generated code before relying on it. Whether the feature changes professional development will depend on how well generated apps hold up beyond simple examples.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...