Google recently quietly releasedAn app called Google AI Edge Gallery, which will allow users to run artificial intelligence models hosted by the Hugging Face service platform on Android phones, and plans to launch an iOS platform version next.
Artificial intelligence models run through the Google AI Edge Gallery can be run offline on the device, eliminating the need for an internet connection. This prevents users' personal information from being uploaded to the cloud and potentially exposed. The app is licensed under the Apache 2.0 license, meaning it can be used without restrictions in a wide range of scenarios, even for commercial purposes.
Currently, the Google AI Edge Gallery is experimental and available in alpha. The app itself is downloaded and installed from GitHub. Once the app is running, the main screen will display specific features that require AI models, such as image recognition and AI chat interaction. Clicking on a feature will display the appropriate AI model.
In addition, Google AI Edge Gallery also provides a feature called Prompt Lab, which allows users to execute artificial intelligence models through commands to handle specific tasks, such as summarizing entire text content or rewriting text content. It can also respond to application modes such as multi-tasking and allows users to fine-tune the model execution results.
Because it operates offline on different mobile phones, Google said that actual performance will vary depending on different hardware and the scale of the artificial intelligence model used.
