Run Google’s new Gemma 4 AI models locally on Android and iOS: Here’s how
To run these AI models, Google has a dedicated app called AI Edge Gallery, where these AI models can be put to task on Android and iOS.
- Apr 6, 2026,
- Updated Apr 6, 2026 10:40 AM IST
Google recently announced Gemma 4, new open AI models designed to bring frontier-level intelligence directly to mobile hardware. These models run locally on smartphones without an internet connection.
Google has a dedicated app called AI Edge Gallery, where these AI models can be put to task on Android and iOS devices. Here’s how users can leverage its reasoning and agentic capabilities.
What is Gemma 4?
Google is expanding its Gemini research capabilities to the open ecosystem with Gemma 4, a new generation of lightweight models built for flexibility.
Unlike closed systems, Gemma is optimised to work across devices without an internet requirement, allowing developers to work across local and cloud environments.
Gemma 4 is available in four variants: Effective 2B (E2B), Effective 4B (E4B), 26B Mixture of Experts (MoE), and 31B Dense.
What is Google AI Edge?
Google AI Edge is a dedicated app that provides a unified development stack, allowing developers to deploy AI models across Android, iOS, and edge devices. This interface allows users to run advanced LLMs and multimodal models, including Gemma 4, fully offline.
Google AI Edge app offers features such as offline functionality that keeps the data private, a conversational AI interface with Thinking Mode, a workspace to generate, summarise, or rewrite text, model Management, and more.
Users can also use Ask Image for visual queries and Audio Scribe for real-time transcription and translation.
How to use Gemma 4 models on Android and iOS via Google AI Edge?
Step 1: Download the Google AI Edge app from the Google Play Store on Android and App Store on iOS. ( For iOS, it is preferred to use iPhone 15 Pro or newer for better performance)
Step 2: Open the app, and select between different modes such as AI Chat, Ask Image, Audio Scribe, or Agent Skills.
Step 3: Now, head towards the Models tab, and select between E2B (fastest) or E4B (most intelligent), based on your device’s capabilities.
Step 4: Once the selected AI model is downloaded, users can run it locally on the device without any major limitations.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
Google recently announced Gemma 4, new open AI models designed to bring frontier-level intelligence directly to mobile hardware. These models run locally on smartphones without an internet connection.
Google has a dedicated app called AI Edge Gallery, where these AI models can be put to task on Android and iOS devices. Here’s how users can leverage its reasoning and agentic capabilities.
What is Gemma 4?
Google is expanding its Gemini research capabilities to the open ecosystem with Gemma 4, a new generation of lightweight models built for flexibility.
Unlike closed systems, Gemma is optimised to work across devices without an internet requirement, allowing developers to work across local and cloud environments.
Gemma 4 is available in four variants: Effective 2B (E2B), Effective 4B (E4B), 26B Mixture of Experts (MoE), and 31B Dense.
What is Google AI Edge?
Google AI Edge is a dedicated app that provides a unified development stack, allowing developers to deploy AI models across Android, iOS, and edge devices. This interface allows users to run advanced LLMs and multimodal models, including Gemma 4, fully offline.
Google AI Edge app offers features such as offline functionality that keeps the data private, a conversational AI interface with Thinking Mode, a workspace to generate, summarise, or rewrite text, model Management, and more.
Users can also use Ask Image for visual queries and Audio Scribe for real-time transcription and translation.
How to use Gemma 4 models on Android and iOS via Google AI Edge?
Step 1: Download the Google AI Edge app from the Google Play Store on Android and App Store on iOS. ( For iOS, it is preferred to use iPhone 15 Pro or newer for better performance)
Step 2: Open the app, and select between different modes such as AI Chat, Ask Image, Audio Scribe, or Agent Skills.
Step 3: Now, head towards the Models tab, and select between E2B (fastest) or E4B (most intelligent), based on your device’s capabilities.
Step 4: Once the selected AI model is downloaded, users can run it locally on the device without any major limitations.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine