Gemma 4 by Google: Revolutionising Mobile AI

Gemma 4 is Google’s latest advancement in open AI models, specifically designed to integrate cutting-edge intelligence into mobile devices. These models function locally on smartphones, eliminating the need for an internet connection.

Understanding Gemma 4

Google is broadening its Gemini research with Gemma 4, introducing a new set of lightweight models which prioritise flexibility. Unlike traditional closed systems, Gemma is specifically developed to operate seamlessly across both local and cloud environments, without necessitating an internet connection. This feature allows developers to create applications that are more versatile.

Gemma 4 is offered in four distinct variants: Effective 2B (E2B), Effective 4B (E4B), 26B Mixture of Experts (MoE), and 31B Dense.

What is Google AI Edge?

Google AI Edge is an innovative application that delivers a comprehensive development environment, enabling developers to implement AI models on Android, iOS, and edge devices. This sophisticated interface facilitates the operation of advanced large language models (LLMs) and multimodal models, including Gemma 4, entirely offline.

The AI Edge app includes functionalities such as offline performance that ensures data privacy, a conversational AI interface featuring Thinking Mode, a workspace designed for text generation, summarisation, or rewriting, as well as model management capabilities and more.

Additionally, users can access Ask Image for visual queries and Audio Scribe for immediate transcription and translation services.

Utilising Gemma 4 Models on Android and iOS via Google AI Edge

Step 1:

Begin by downloading the Google AI Edge app from the Google Play Store for Android devices or the App Store for iOS. It is advisable to use an iPhone 15 Pro or a later model for optimal performance on iOS.

Step 2:

Launch the app and choose from various options such as AI Chat, Ask Image, Audio Scribe, or Agent Skills.

Step 3:

Navigate to the Models tab and select E2B for speed or E4B for enhanced intelligence, based on the specific capabilities of your device.

Step 4:

After downloading the chosen AI model, users can execute it locally on their device without facing significant limitations.