Google's New AI App: Run AI Models Locally

Google Launches AI Edge Gallery for On-Device AI Processing

Recently, Google has introduced an application enabling users to execute a variety of open-source AI models directly on their mobile devices.

The application, known as Google AI Edge Gallery, is currently available for Android, with an iOS version planned for future release. It facilitates the discovery, download, and execution of AI models capable of image generation, question answering, code creation, and text editing, among other functions.

Benefits of On-Device AI

While cloud-based AI models often offer greater processing power, they also present certain drawbacks. Concerns regarding data privacy and the necessity of a stable internet connection are key considerations for many users.

Running models locally addresses these issues by processing data directly on the device, eliminating the need to transmit information to remote servers and ensuring functionality even without network access.

Exploring the Google AI Edge Gallery

Google describes the AI Edge Gallery as an “experimental Alpha release” and it can be obtained via GitHub following the provided instructions.

The app’s interface features shortcuts to various AI functionalities, including “Ask Image” and “AI Chat.” Selecting a function displays a curated list of suitable models, such as Google’s Gemma 3n.

Features and Functionality

The Google AI Edge Gallery incorporates a “Prompt Lab” designed for initiating “single-turn” tasks powered by the installed AI models.

This lab provides pre-defined task templates and adjustable settings, allowing users to customize model behavior for tasks like text summarization and rewriting.

Performance Considerations

Performance levels will vary depending on the device’s capabilities, Google notes.

More powerful hardware will naturally result in faster model execution, but the size of the AI model itself also plays a significant role; larger models require more processing time.

Open-Source and Community Driven

Google is actively seeking feedback from the developer community to enhance the Google AI Edge Gallery experience.

The application is licensed under Apache 2.0, granting broad usage rights – including commercial applications – without restrictions.

• Key Feature: Offline AI processing
• Supported Platforms: Android (iOS coming soon)
• License: Apache 2.0