Explore Android AI Samples: Gemini, Imagen & More

What is the Android AI Sample Catalog?

The Android AI Sample Catalog is a stand‑alone application curated by Google’s Android team that bundles dozens of self‑contained demos. Each sample demonstrates a specific AI capability—ranging from text‑to‑image generation with Imagen to Gemini‑powered video summarization—using the latest Google Cloud and on‑device GenAI APIs.

Unlike typical research notebooks, these samples focus on clear, production‑style code. They are written in Kotlin, use Gradle for dependency management, and are fully compatible with the current Android Studio IDE. They provide ready‑to‑run, minimal‑setup examples that make it easy for Android developers to try generative AI in their apps.

Why it matters: If you’re building an Android app that needs image generation, speech‑to‑text, or on‑device summarization, the catalog gives you a quick, reliable reference. You can copy the sample code, tweak the model parameters, and integrate the same patterns into your own project.

How to Get Started

1. Clone the repo

   git clone https://github.com/android/ai-samples.git
   

2. Open in Android Studio – the repo contains a full app module and a samples/ folder that each house individual examples.
3. Set up Firebase – many Gemini‑powered samples require a Firebase configuration (google‑services.json). Follow the https://developer.android.com/ai/sample-firebase guide to add the file under /app.
4. Sync & Run – build the project, choose a target device, and the catalog’s Sample List screen will appear.

Each sample has its own README and screenshots, making it straightforward to identify the feature you want to experiment with.

Highlights of the Sample Catalog

Sample	Description	Key Tech	Screenshot
Gemini Image Chat	Converse with Gemini 3 Pro Image model to generate and tweak images.	Gemini 3, LLM
Gemini Chatbot	Build a chatbot using Gemini Flash; adjust system prompts to change tone.	Gemini Flash, LLM
Gemini Multimodal	Submit an image to Gemini Flash for text and image‑to‑text responses.	Gemini Flash, Multimodal
Imagen Generation	Generate landscape or portrait images in various artistic styles.	Imagen
Magic Selfie	Remove background with ML Kit subject segmentation and replace it using Imagen.	ML Kit, Imagen
Gemini Video Summarization	Use Gemini Flash to produce a concise recap of a long video.	Gemini Flash, Video
On‑Device Summarization	Summarize large texts locally with Gemini Nano.	Gemini Nano, GenAI
On‑Device Writing Assistance	Proofread and rewrite text using Gemini Nano Rewriting API.	Gemini Nano, Rewriting
On‑Device Image Description	Generate descriptive captions for images without network calls.	Gemini Nano, Image Description

Note: The catalog is ever‑evolving. New samples are published regularly, and the Android team is actively expanding the range of capabilities.

How to Contribute

While the project currently does not accept external contributions, developers can fork the repo, extend or refactor existing samples, and share their improvements with the community. All changes should comply with the Apache‑2.0 license under which the samples are distributed.

License & Legal

The repository is licensed under the Apache License 2.0. Feel free to use the code in your own apps, provided you adhere to the license terms.

Take the Next Step

1. Browse the catalog and pick a sample that fits your use case.
2. Modify the sample to suit your app’s UI/UX.
3. Leverage the same Google AI SDKs to add new generative AI features.
4. If you discover bugs or have feature requests, open an issue on the GitHub repo.

Final Thought

The Android AI Sample Catalog is more than just a set of demo apps; it’s a toolbox that accelerates AI experimentation on Android. Whether you’re exploring image generation, on‑device summarization, or intelligent voice interfaces, the catalog makes it simple to prototype, iterate, and deploy high‑quality generative‑AI features in your next app. Happy coding!