Gemini in Chrome is coming to Android, and it does more than summarize a page. Google says Chrome on Android will get a Gemini-powered browsing assistant, image customization through Nano Banana, and an agentic "auto browse" feature that can handle selected web tasks with user confirmation.
The first rollout is narrow. Google says Gemini in Chrome starts at the end of June for select Android devices in the United States, with Android 12 or higher, English-US language settings and at least 4GB of RAM. Auto browse starts even narrower, with AI Pro and Ultra subscribers in the U.S. on select eligible devices.
What Gemini in Chrome does
Gemini in Chrome reads the page you're viewing, so you don't have to paste text into a separate chatbot. Google says you'll tap the Gemini icon in Chrome's toolbar and ask questions about the current page.
On a phone that covers a lot of small jobs:
- summarizing long articles and explaining complex topics in simpler language
- asking questions about a page
- pulling recipe items into Keep and adding event details to Calendar
- finding information from connected Google apps
- turning page content into a more visual format
The difference from a blank assistant prompt is context. You ask about the thing already on screen.
Rollout details
| Feature | Initial rollout |
|---|---|
| Gemini in Chrome on Android | End of June 2026 |
| Region | United States |
| Device requirement | Select Android devices with Android 12 or higher |
| Memory requirement | 4GB RAM or more |
| Language | English-US |
| Auto browse | AI Pro and Ultra subscribers in the U.S. on select eligible devices |
Don't expect a normal Play Store update. If your phone is outside the U.S., set to another language, below the RAM requirement or not in the first device group, you may not see the feature right away. Chrome version, account eligibility and regional rollout can all affect access, so Android version alone won't tell you.
Auto browse
Auto browse is Google's agentic browsing feature for Chrome on Android. Instead of only answering questions, Chrome can take steps across a website or supported service to finish a task. Google's examples include reserving parking using details from a ticket confirmation and updating a recurring pet-food order.
It's designed to ask for confirmation before sensitive actions like making purchases or posting on social media. That step is the line between helping you navigate a task and the browser deciding for you. Agentic browsing will only feel trustworthy if you can see what it's doing, pause it and approve the final action.
Nano Banana image tools
Google is also bringing Nano Banana image customization into Chrome on Android. You can ask the browsing assistant to create or modify visuals based on web content. Google's examples are turning a study page into an infographic and changing an apartment-listing image to show a furnished room.
| Use case | How it helps |
|---|---|
| Studying | Turn dense text into visual study notes |
| Shopping | Compare product details or visualize changes |
| Travel | Summarize pages and plan from bookings |
| Work research | Extract key points from articles and documents |
| Design ideas | Modify web images for rough exploration |
AI-edited images are fine for planning and ideas. Don't mistake them for verified product photos, legal documents or medical information.
What changes on mobile
Long tasks are awkward on a phone. Screens are small, tabs pile up, and copying data between pages, notes, calendars, carts and email is tedious. Gemini in Chrome attacks that in three steps. It explains what's already on the page, whether that's an article, a product page, a support doc or a complicated form. It connects to Google apps, which matters because many tasks end in Calendar, Keep, Gmail, Maps or a shopping cart rather than on a web page. And auto browse moves from advice to action.
The browser also sees sensitive pages: banking, health, personal email, orders, travel and work tools. Google says the feature includes protections against emerging threats like prompt injection, but be selective about where you turn AI assistance on.
Settings to check first
- Account. Chrome, Gemini, Gmail, Calendar and Keep can involve different Google accounts. Make sure the assistant is connected to the one where you want help.
- Personal Intelligence. Google says some personalized responses depend on opting in. If you don't want Gemini using connected app context, leave it off or review the controls.
- Confirmation screens. For auto browse, read what Chrome is about to do before you tap through, especially for purchases, bookings, uploads, messages and public posts.
- Sensitive sites. On banking, legal, medical, workplace or identity pages, don't ask the assistant to handle anything you don't fully understand.
- Chrome updates. Google is stressing security protections, so stay on current Chrome builds.
Who benefits most
People who do real work from their phone: students, creators, frequent travelers, shoppers, small-business owners and anyone who reads long pages on mobile. It helps most if you already use Gmail for confirmations, Calendar for events, Keep for lists, and Google Photos or web images for visual planning. If your workflow lives mostly outside Google services, the page summaries and Q&A still work, but the task automation may feel thinner at launch.
Gemini in Chrome vs the Gemini app
The Gemini app starts from a prompt. Gemini in Chrome starts from the current page.
| Question | Gemini app | Gemini in Chrome |
|---|---|---|
| Best starting point | A general question | A page you are already viewing |
| Context | You provide it | Chrome can use page context |
| App connections | Depends on settings | Designed around browsing and Google apps |
| Task automation | Broader assistant tasks | Web task flow through auto browse |
| Best use | Chat, drafting, planning | Reading, researching, acting on web pages |
Use the app for open-ended brainstorming and Chrome when the source material is already in front of you.
When to turn it on
If you're eligible and comfortable with Google-connected AI features, test it when it arrives. Start with low-risk tasks like summarizing articles, explaining support pages, saving recipe ingredients or asking about a product comparison.
Hold off on auto browse for anything high-stakes. Booking, ordering, payments, work tools and public posting all deserve slower trust. Give it simple tasks first and widen from there.
FAQ
When is Gemini in Chrome coming to Android?
Google says it starts rolling out to eligible Android users at the end of June 2026.
Which Android phones get it first?
Select Android devices in the U.S. running Android 12 or higher, with at least 4GB of RAM and English-US language settings.
Do I need a paid Google AI plan for auto browse?
Yes. Google says auto browse starts for AI Pro and Ultra subscribers in the U.S. on select eligible Android devices.
What is Nano Banana in Chrome?
Google's image customization capability inside Chrome. It lets you create or alter visuals based on page context and prompts.
