Google Gemini-M3P adds UI prototyping API for Android Studio

Tech · 5 min read

The new Gemini-M3P prototyping API helps Android developers convert design artifacts into runnable prototypes inside Android Studio. It supports Figma imports, hand-drawn sketches, and plain language prompts, and outputs Compose UI code with appropriate semantics and TalkBack labels. Google highlights its attention to platform conventions like navigation patterns, form controls, and system spacing.

Gemini-M3P also simulates device APIs and state transitions so prototypes can include camera access, map interactions, and low-level system callbacks. The API includes a preview layer that replicates battery and connectivity conditions for realistic usability testing. Developers can flag generated components as review required, triggering a suggested changelist for human oversight.

Privacy controls allow models to run in Google Cloud or on private Vertex AI instances where enterprises can manage data retention. Early adopters in fintech and healthcare praised the speed of prototyping but requested more robust support for legacy view hierarchies and backward compatibility for older Android versions.