<ul data-eligibleForWebStory="true"><li>Google has launched Gemma 3N, an advanced AI model designed for smartphones and edge devices, promising high performance without compromising intelligence.</li><li>Gemma 3N is a compact, multilingual, multimodal AI model that can run on devices with just 2-3 GB of VRAM, making it suitable for consumer-grade devices.</li><li>This model achieved an E4B score, making it the first sub-10B model to reach this milestone, and it incorporates higher transformer layers for improved performance in chat and long-form generation.</li><li>Gemma's visual understanding and chain-of-thought prompting enhance accuracy on reasoning-heavy audio tasks, providing advanced tools for building local assistants or edge agents.</li></ul>

Google Launches Gemma 3N: The Most Powerful Open-Source AI Model for Phones & Edge Devices

Discover more