Last week, Google quietly released an application that allows users to launch a series of opening AI models from the AI Dev platform, hugging their face on their phones.
Called Google AI Edge galleryThe application is available on Android and will soon come on iOS. It allows users to find, download and run compatible models that generate images, answer questions, write and edit the code and many others. The models work offline, without the need for internet connection, using supported phone processors.
AI models operating in the cloud are often stronger than their local counterparts, but also have their drawbacks. Some users may be careful about sending personal or confidential data to a remote data center or wanted to have available models without having to find Wi-Fi or a cellular connection.
Google Ai Edge Gallery, which Google calls “Experimental Alpha Release”, can be downloaded from GitHub, watching these instructions. The main screen shows shortcuts for AI tasks and possibilities such as “Ask Image” and “Ai Chat”. Touching the function pulls out a list of models adapted to the task, such as Google Gemma 3N.
The Google AI Edge gallery also provides a “fast laboratory”, which users can use to start “disposable” tasks powered by models, such as summary and rewriting text. The assembly laboratory is equipped with several task templates and configurable settings to tune the behavior of models.
Your course may differ in terms of performance, it warns Google. Modern devices with stronger equipment will be predictably launched by models faster, but the size of the model also matters. Larger models will take more time to complete the task – say, answering the image question – than smaller models.
Google programmers inviting members of the community to provide feedback on the Google AI Edge gallery. The application is under the APACHE 2.0 license, which means that it can be used in most contexts – commercial or otherwise – without restrictions.
