Android Studio adds ‘agentic AI’ with Journeys and Agent Mode

At its Google I/O 2025 developer conference on Tuesday, Google announced that Android Studio—the official IDE for Android development—is getting a major AI enhancement.

Alongside the release of the latest Gemini 2.5 Pro model, Android Studio is introducing a new “agentic AI” feature called Journeys and an upcoming “Agent Mode” to support more advanced development tasks.

Natural Language App Testing with Journeys Powered by Gemini

Journeys, powered by Gemini, enables developers to test their apps by describing user flows and expected outcomes in natural language. Gemini then automatically runs these tests.

Google’s feature aims to simplify app testing for developers by reducing the need for extensive code. Though experimental, it seeks to speed up high-quality code delivery by streamlining testing, validation, and issue reproduction.

Seamless Testing Across Devices with Results in the IDE

Google adds that tests can be executed on both physical and virtual Android devices, with results displayed directly within the IDE.

Another upcoming feature is Agent Mode, an autonomous AI capability powered by Gemini. This tool is designed to assist developers with more complex, multi-step development tasks. For example, when integrating a new API, the agent can generate an execution plan, add necessary dependencies, modify files, and resolve bugs.

Additional AI enhancements in Android Studio include an upgraded “crash insights” feature in the App Quality Insights panel. Now powered by Gemini, it can analyze app crashes, identify potential causes in the source code, and suggest fixes.

Additionally, Google is introducing a new “Studio Labs” option in Android Studio’s Settings menu, allowing developers to experiment with its AI features, which are still in development. This option will be available only in stable versions, beginning with the release codenamed Narwhal.

Another new experiment now available is the public preview of Android Studio Cloud, which can be accessed through Firebase Studio. This service streams a Linux-based instance of Android Studio directly to your web browser, allowing developers to build Android apps from anywhere with an internet connection.

Coming soon, the Version Upgrade Agent, powered by Gemini, will assist in automating dependency updates within projects.

Enhanced Developer Productivity with Gemini’s Intelligent Code and UI Tools

Gemini will also offer several productivity-boosting capabilities, including automatic generation of Jetpack Compose preview code, UI code transformation in the Compose Preview using natural language, the ability to attach image files (such as mockups or screenshots) to AI prompts, and the option to include project files as context in Gemini chats. Additionally, a new “Rules in Gemini” feature will let developers customize coding styles and output formats.

Google is also introducing an enterprise-grade version of its AI-enhanced Android Studio, called Gemini in Android Studio for businesses.This offering lets organizations securely access AI features by subscribing to Gemini Code Assist in Standard or Enterprise editions, ensuring data protection.

New Features and Enhancements

Additional updates include resizable previews, improved Compose Preview navigation, a default embedded Android XR emulator, and enhancements to Backup and Restore, and Backup and Sync. There are also a few enhancements coming to Android’s Kotlin Multiplatform support.

Google also announced it will support developers in adapting to Android’s upcoming shift to 16KB memory page sizes—a fundamental architectural change—by providing early warnings and tools to test apps in the new environment.

Read the original article on: Techcrunch

Read more: Google Is Integrating Gemini into Android Auto for In-Car Use