Last week, Google quietly launched a new app that allows users to run a variety of open-source AI models from the Hugging Face platform directly on their smartphones.

Called Google AI Edge Gallery, the app is currently available for Android, with an iOS version on the way. It lets users search for, download, and run compatible models that can perform tasks like generating images, answering questions, and writing or editing code — all without needing an internet connection. The app leverages the processing power of supported mobile devices to run these models locally.

While cloud-based AI models are generally more powerful, they come with trade-offs. Users may be concerned about sending personal or sensitive data to remote servers or may want to access AI tools without relying on Wi-Fi or mobile data.

Google describes the app as an “experimental Alpha release,” available for download via GitHub with installation instructions provided. The app’s home screen offers shortcuts to various AI tasks and features, such as “AI Chat” and “Ask Image.” Selecting a task brings up a list of suitable models, including Google’s own Gemma 3n.

One feature, called Prompt Lab, enables users to run “single-turn” AI tasks like summarizing or rewriting text. It includes pre-built templates and customizable settings to help users fine-tune how the models behave.

Google notes that performance may vary. Newer phones with more powerful processors will handle models more efficiently. However, larger models will generally take longer to complete tasks, like analyzing an image or answering a question.

The company is encouraging developers to test the app and provide feedback. Released under the Apache 2.0 license, the app can be freely used in both personal and commercial projects.