Android Studio, the built-in growth setting (IDE) for Android app builders, is getting an AI improve, Google introduced at its developer convention, Google I/O 2025, on Tuesday.
Along with the rollout of the newest Gemini 2.5 Professional mannequin, Android Studio is gaining a brand new “agentic AI” functionality known as Journeys and can quickly introduce an “Agent Mode” for extra advanced growth duties.
Utilizing Gemini, journeys will enable builders to check their app by describing the actions and assertions in pure language for the consumer journeys throughout the app. Gemini will then carry out the exams for you.
The characteristic, explains Google, will let builders check their apps extra simply, with out having to put in writing in depth code to take action. The corporate cautioned that that is nonetheless experimental, however finally, the purpose is to extend the velocity of delivery high-quality code whereas lowering the time it takes to check, validate, or reproduce points.
The exams can run on bodily or digital Android gadgets, and their outcomes will seem instantly within the IDE, Google says.
One other soon-to-arrive addition entails an autonomous AI characteristic powered by Gemini known as Agent Mode. This may enable builders to make use of varied instruments to deal with extra advanced, multi-stage growth duties. As an illustration, if a developer is attempting to combine a brand new API, the agent might give you an execution plan that provides the mandatory dependencies, edits recordsdata, and fixes bugs.
Different AI options coming to Android Studio embody a Gemini-powered enchancment to the App High quality Insights panel’s “crash insights” characteristic, which might now use AI to assist decide what in an app’s supply code might have triggered the app to crash, and recommend a repair.
Plus, Google will now enable builders to check out its nonetheless experimental AI options via a brand new “Studio Labs” menu within the Settings menu of Android Studio. This feature shall be out there in steady releases solely, beginning with the discharge codenamed Narwhal.
One other experiment now out there is the general public preview of Android Studio Cloud. Accessed via Firebase Studio, the brand new service streams a Linux machine working Android Studio to your internet browser, enabling Android growth wherever you will have entry to an web connection.
A Model Improve Agent will quickly arrive as a part of Gemini in Android Studio to assist automate dependency upgrades.
Gemini will even assist builders routinely generate Jetpack Compose preview code, rework UI code inside the Compose Preview setting utilizing pure language, connect picture recordsdata (like UI mockups or screenshots) to AI prompts, connect mission recordsdata as context in chats with Gemini, and arrange most well-liked coding types or output codecs with a brand new “Rules in Gemini” characteristic.
The corporate can also be rolling out an enterprise-ready model of its AI-powered Android Studio with the launch of Gemini in Android Studio for companies, which lets groups deploy AI whereas preserving information protected when subscribing to Gemini Code Help in Customary or Enterprise editions, it says.
Different updates embody resizable previews in Compose Preview and navigation enhancements, an embedded Android XR emulator that launches by default within the embedded state, and upgrades to Backup and Restore and Backup and Sync, amongst different issues. Android’s Kotlin Multiplatform will even see a small handful of enhancements.
As well as, Google says it is going to assist builders put together for Android’s 16KB web page sizes — a change to Android’s underlying structure — with early warnings and instruments for testing apps within the new setting.
