Google is building Gemini Nano AI right into Chrome

Google is building Gemini Nano AI right into Chrome


An image showing the Chrome logo surrounded by yellow circles
Image: The Verge

Google is building its Gemini AI into Chrome on desktop. During its I/O event on Tuesday, Google announced that Chrome 126 will use Gemini Nano to power on-device AI features, such as text generation.

Gemini Nano is the lightweight large language model Google introduced to the Pixel 8 Pro last year — and, later, the Pixel 8. To get Gemini Nano on Chrome, Google says it tweaked the model and optimized the browser to “load the model quickly.”

The integration will let you do things like generate product reviews, social media posts, and other blurbs directly within Chrome. Microsoft similarly added its AI assistant Copilot to Edge last year, letting you ask questions and summarize the information on your screen. Unlike Gemini Nano in Chrome, Copilot in Edge doesn’t run locally on your device.

Google also announced that it will make Gemini available in Chrome DevTools, which developers use to debug and tune their apps. Gemini can provide explanations for error messages, as well as suggestions on how to fix coding issues.