Google is bringing generative AI straight to your smartphone.
In late May 2025, the tech giant launched the AI Edge Gallery, an Android app that enables users to download and run AI models directly on their devices, which eliminates the need for a cloud connection. While the app is currently only available for Android users at the time of this writing, Google plans to expand its marketplace to include iOS users in the near future.
According to the GitHub repository for Google’s AI Edge Gallery: “The Google AI Edge Gallery is an experimental app that puts the power of cutting-edge generative AI models directly into your hands.”
Understanding the key features
Currently available, the Google AI Edge Gallery features several key components that help users maximize their experience with on-device AI. Some of the most useful features include:
- AI Chat: The typical AI chatbot experience that most of us are already familiar with.
- Ask Image: Upload a digital image and ask the AI specific questions about it.
- Prompt Lab: Use freeform prompts to try out various single-turn LLMs interactions or generate, rewrite, and summarize code.
All three features are available from the home screen of the Google AI Edge Gallery. After selecting one, you’ll be asked to pick from one of the compatible AI models that are available through the marketplace. If the AI model isn’t already on your smartphone, you’ll be given the option to download and install it.
SEE: AI Quick Glossary from TechRepublic Premium
More Google coverage
- New Google Search AI Mode is ‘Total Reimagining,’ Says CEO Sundar Pichai
- In Major Ruling, Judge Finds Google ‘Willfully Acquired and Maintained Monopoly Power’ Over Digital Ad Market
- Google’s Big Bet on Nuclear Energy: ‘The Race to Power AI-Driven Data Centers is Accelerating’
- Computer History Museum Releases Original AlexNet Code: Why It Matters
Exploring available AI models
The Google AI Edge Gallery currently hosts 18 AI models. This number is expected to grow as smartphones become more powerful and as more AI model developers begin to embrace mobile platforms.
It’s no surprise that the majority of AI models featured in the Google AI Edge Gallery were developed by various teams at Google; however, there are other models available as well including some from industry giants such as Microsoft, and a few from lesser-known brands.
- Google: MedGemma-27B, Gemma3-1B, Gemma3-27B, Gemma3-12B, Gemma3-4B, Gemma2-2B, MediaPipe, Gecko-110m
- Microsoft: Phi-4-mini
- DeepSeek Technologies: DeepSeek-R1
- Alibaba Cloud: Qwen2.5-0.5B, Qwen 2.5-3B, Qwen 2.5-1.5B
- Ollama: TinyLlama-1.1B, Llama-3.2-3B, Llama-3.2-1B
- MadeAgents: Hammer2.1-1.5b
- Hugging Face: SmolLM-135M
Paving the way for AI on the go
Although the concept of on-device AI isn’t exactly unheard of, it remains a rarity among modern AI models.
Most consumer phones lack the processing power to run some of the larger and more sophisticated models, as seen in the limited options of Google’s AI Edge Gallery. Still, the marketplace represents a significant step forward in bridging the gap between smartphones and AI.