Integrate your app with the Android intelligence system

Android is evolving into an intelligence system, powered by AI. Integrate your app through a self-service model, and become part of the next generation of intelligent experiences on Android.
AI on Android

Find your Android AI pathway

Select from Google's cutting-edge AI/ML solutions to empower and enlighten your Android app users.
Make your app's capabilities available to qualified agents using AppFunctions or rely on UI automation on supported devices as a fall back.
Implement on-device features with Gemini Nano and ML Kit GenAI APIs, or harness the full power of Cloud and Hybrid AI with Firebase AI Logic.
Leverage agentic tools for Android development, including Gemini in Android Studio, Android CLI, and the official suite of Android skills, to accelerate your workflow.

New Android intelligent features

New feature
AppFunctions is an Android API designed to simplify building Android MCP integrations. It empowers your apps to behave like on-device MCP servers, contributing functions that act as tools for use by proactive features along with client agents and assistants. AppFunctions integration with Gemini is currently in a private preview with trusted testers, but you can begin preparing your apps now using AppFunctions and development tools.
New feature
Prepare your app for automation with Android Computer Control. This framework lets OEM-preloaded assistants interact with supported apps using task automation, with no code changes required by your app.
New feature
Take advantage of Gemini Nano's audio capabilities with the Speech Recognition API, and reduce inference times when using Prompt API with the prefix caching feature.
New feature
Join the AICore Developer Preview to prototype with on-device Gemini Nano models. Test custom prompts and optimize performance.
New feature
Unlock local agentic intelligence on Android with Gemma 4, our open model featuring advanced reasoning and tool calling. Code with Gemma 4 in Android Studio, and build on-device AI features with ML Kit GenAI APIs today.

AI case studies

By implementing Gemini Nano on-device, Kakao Mobility streamlined address entry and reduced order completion time by 24% while reducing server costs and enhancing user privacy.
Read about how Android developers are using Firebase AI to add generative AI features to their apps.
Gemini Nano with multimodality enables Talkback to provide clear and detailed image descriptions even if the device is offline or on an unstable network connection.
Google Pixel's voice recorder app leverages Gemini Nano to provide offline and on-device inference to generate summaries of voice recordings.

Traditional ML solutions

Harness the processing capabilities of Android devices for on-device ML detection and classification of images, audio, and text.
ML Kit provides production-ready solutions to common tasks and requires no ML expertise. Models are built-in and optimized for mobile. ML Kit is easy to use and lets you focus on feature development rather than model training and optimization.
For more control, or to deploy your own ML models, Android provides a custom ML stack built on top of LiteRT and Google Play services that covers the essentials needed to deploy high performance ML features.
MediaPipe offers open source cross-platform, customizable ML solutions for live and streaming media.
Deploy and manage your custom ML and GenAI models for on-device AI features efficiently with Play for On-device AI. Google Play simplifies model delivery, helping you to enhance user experience while keeping app size optimized.

Accelerate your development

AI-powered tools can automate repetitive tasks, suggest code, debug errors, and more—which lets you focus on higher-level problem-solving and innovation.
Gemini in Android Studio is your coding companion for Android development. It can generate code, find relevant resources, learn best practices, troubleshoot errors, and save time.
Android CLI makes it faster and easier to build Android apps using the agents and developer environments of your choice.
Get early access to the latest Gemini Nano models. Test your prompts, explore on-device performance, and help shape the future of on-device AI before it hits production.
The fastest way to start using Gemini is with Google AI Studio, a web-based tool that lets you prototype, run prompts right in your browser, and get started with the Gemini API.

Latest news and videos

Community & Support

If you are building generative AI features, we would love to have a conversation with you! Complete the form to keep in touch, or join an existing community group.