Run powerful open-source AI models directly on your mobile device. Experience generative AI offline, privately, and with high performance. Explore features like agent skills, AI chat with reasoning visualization, and image/audio analysis.
1.0.7
June 22, 2026
- Add support for importing LiteRT-LM models via Hugging Face community URLs.
Google AI Edge Gallery is a free application that allows you to run powerful open-source Large Language Models (LLMs) directly on your mobile device. It provides a private, offline, and high-performance experience for generative AI.
No, Google AI Edge Gallery does not contain advertisements. It is a completely free application with no ads or in-app purchases.
The app supports a wide variety of open-source models, including official support for the Gemma 4 family. You can easily download models from a list or load your own custom models.
All model inferences happen directly on your device hardware, meaning no internet connection is required for processing. This ensures total privacy for your prompts, images, and sensitive data.
Agent Skills transform your LLM into a proactive assistant by augmenting its capabilities with tools like Wikipedia, interactive maps, and visual summary cards. You can load modular skills from a URL or browse community contributions.
Yes, the 'Thinking Mode' in AI Chat allows you to see the model's step-by-step reasoning process, which is especially useful for understanding complex problem-solving. This feature currently works with supported models like the Gemma 4 family.
Google AI Edge Gallery is currently supported on iPhones. Performance is dependent on your device's hardware (CPU/GPU).
The app is in active development. The latest version is 1.0.2, and it was last updated on April 3, 2026, indicating a commitment to ongoing improvements and feature additions.
