Source: developers.googleblog.com/build-full-stack-android-apps-with-gemini-in-google-ai-studio
The Build agent detects when a prompt needs persistent storage, user sign-in or generative AI, and with the user's confirmation it provisions a Cloud Firestore database, configures Firebase Authentication with Google Sign-In, and generates per-user Security Rules. Calls to Gemini models are routed through Firebase AI Logic with Firebase App Check, so API keys are not embedded in the APK.
The in-browser Android emulator now starts with lower latency, can switch its stream to 1200p HD, and plays real-time audio in the browser. Developers can click Take a screenshot in the emulator toolbar to attach the live screen to a prompt for visual UI changes. The agent also uses the official Android CLI for project scaffolding and Gradle builds.
Apps can be published from the publish panel to a personal Google Play library using a standard Google account, and AI Studio builds and signs the Android App Bundle. The features are rolling out in Google AI Studio starting today, and the text says testing on your own device through the Play Store costs nothing.
Why it matters if you build agents
Agent builders can prototype mobile front ends that store user data in Firestore and call Gemini without shipping API keys in the app. They still need to review the Security Rules the agent generates for their data model.
Teams should still test on a physical phone, since the announcement notes that emulator testing does not replace it.
Get stories like this every weekday
One short email with the day's agent tooling news.