How to use Gemini-Powered Auto Browse Feature on Google Chrome

Gemini in Google Chrome was first launched in September 2025 as a side panel integration in the browser that could answer questions about webpages. However, it was not until January that the feature really began gaining popularity. The reason? The new Auto Browse feature. Similar to Perplexity’s Comet or the Dia browser, it can perform tasks such as booking tickets and making a purchase autonomously by taking control of the browser temporarily. Gemini in Chrome is currently only available in the US, but is expected to be expanded to more regions soon.

What Is Auto Browse in Google Chrome

Auto Browse is an AI-assisted browsing feature powered by Google’s Gemini models that allows Chrome to complete multi-step tasks on the web. Rather than simply summarising information, the system can actively navigate websites, search for information, compare options, and carry out actions like filling forms or placing items in a shopping cart.

The feature is designed to work through a dedicated Gemini side panel in Chrome, where users can type natural language instructions. From there, the AI agent can perform actions such as researching travel plans, comparing products, or organising information across multiple web pages.

Who Can Use Auto Browse in Chrome?

At the time of writing, Auto Browse is being rolled out gradually and is currently available to Google AI Pro and Ultra subscribers in select regions, including the US. It is expected to be expanded to more regions soon. However, it will likely not be available to those on the free tier of Gemini.

How to Enable and Use Auto Browse in Google Chrome

The setup takes less than a minute, as long as you have the required subscription.

1. Open Chrome on your desktop or laptop.
2. Look for the Gemini sparkle icon in the top-right toolbar.
3. Click it to open the side panel on the right side of the browser.
4. If Auto Browse is available in your account, you will see an option or prompt inside the panel.
5. Type your request in natural language in the text box.
6. Hit submit and watch the agent get to work. A small Auto Browse icon appears on the active tab to show it is running.

During the task, Chrome may display automated clicks, scrolling actions or page navigation as the system completes the request. Some tasks may also require user approval before actions such as purchases or sign-ins are executed. The user always remains in control and can steer the task in a different direction by submitting additional prompts. They can also pause the action or take control back. The AI agent will not take any irreversible steps without asking for permission first.