On Sunday, Google unveiled the Google AI Edge Gallery, an innovative Android application that allows users to operate artificial intelligence (AI) models directly on their devices. Developed by the tech giant based in Mountain View, this experimental app facilitates the discovery, downloading, and execution of large language models (LLMs) without the need for an Internet connection. Currently, it offers a limited range of functionalities, such as chatting, image analysis, and various text and coding operations. Additionally, the company has announced plans to release a version for iOS shortly.
Google AI Edge Gallery App Is Available on GitHub
The app can be accessed through an “experimental alpha release” on Google’s GitHub repository. Users can download an Android application package (APK) file along with comprehensive installation instructions. The file size is 115MB, and it is offered under a permissive Apache 2.0 license, permitting both academic and commercial use.
Staff members from Gadgets 360 successfully downloaded and installed the app without issues. For those unfamiliar with APK installations, Google provides a detailed guide. However, it’s important to exercise caution when downloading APK files, as these apps are not verified by mainstream app marketplaces and could potentially harbor malware. Fortunately, no malware or viruses were detected in this app during testing.
The Google AI Edge Gallery application features a variety of AI models that users can download and utilize locally. The selection of available models may vary based on the device’s specifications and whether it is equipped with an AI-capable chipset. Additionally, users have the option to import and operate models that are pre-installed on their devices.
Three main features are highlighted in the app’s functionality. The first, Ask Image, enables users to perform image analysis. After downloading and launching a model, users can upload an image and query the AI about its content. The second feature, AI Chat, allows users to engage in conversational interactions with the model. However, because the app is limited to local models, some information may not be current.
The third feature, Prompt Lab, offers a suite of AI-driven tools, including tone-based rewriting, text summarization, free-form generation, and code snippet creation. Users have the ability to adjust various parameters such as tokens, temperature, and accelerators while running any AI model. They can also monitor the performance metrics associated with the models.
