Announcing big changes to Gemini in Chrome - agentic browsing with Auto-browse, Nano Banana & more! 🚀

Here's what's new in Gemini in Chrome: 🚀 Agentic Auto-Browsing which can handle both complex and routine tasks. ​🍌 Integration with Nano Banana for editing and transforming images directly 👨‍🏭 ​Enhanced Google Workspace Integration with deeper connectivity to apps like Gmail, in-line editing and more and the new UX gives you a streamlined sidebar which lets you chat and pull context across tabs. Super powerful.

Gemini in Chrome is rolling out to more people in more languages soon. It can work across tabs, schedule meetings in Calendar, send emails in GMail and understand YouTube videos.

looks incredible. cant wait to use it instead of the claude extension which im pretty disapointed at

Looking forward to hearing what you think when you get a chance to try it 🙏

@AnerCohen Is it available now?

@AnerCohen Auto browse is currently in-progress rolling out

hopefully part of the plan is an SDK us builders can use so we can make our web apps Gemini extension-friendly, like an MCP

A few folks have asked how to make their experiences agent-browser friendly. We'll definitely be looking at publishing more guidance and tooling here.

wow thanks Addy, i could imagine the browser agent doing authenticated calls from our users' prompt then your agent is generating UI for analytics or reports on the fly

Is this available outside the US?

We typically roll out Gemini features to the US first and then explore a global release. Looking forward to more of the world getting access!

Very cool Docket @ThisWeeknAI

@ThisWeeknAI Thanks, Jason! 🙏

So excited for this! Gooogle, let's go!

@Workflow_AI 🚀🚀🚀

impressive updates

Thanks Scott!

🤯

🙏

@addyosmani this is amazing, I can't wait to try it. It's not working for me yet, Ultra subscription (Workspace account) and Chrome updated... why not?

The US rollout is happening now so it should be with you in the coming days!

Oh, I thought this was it, cool, thank you! 🙏🏻

@addyosmani it's been a few days but either I don't have the correct Chrome version or this doesn't work. I did your query *exactly* but it just searches Google in the side pane. It does not take over the Chrome window like your demo. Screenshots below:

Can this automate reptitive task such as filling out customer info on the crm? That would be fire.

Form filling is one of the use-cases supported, yep.

There should be a headless or background mode, where it would do this research and come back to me, otherwise it's me staring on a screen

Totally agree and raised this with the team early. I know they're thinking about the right UX and security model for more background tasks.

Does it have apis so we can call it to do things?

How do websites block this shit?

Dude this is siiick, will cover tomorrow on @thursdai_pod 🫡 🙏

@thursdai_pod Excited to hear what you think of it!

Firms are OBSESSED with solving and demonstrating the travel booking problem yet it’s not as big a problem as they think. I don’t need help with my travel. I suspect few do.

Can’t wait to see how Auto-browse actually work in action!

Looking forward to hearing what folks think!

Available in EU in 2029 😭

Are users in Europe also Getting this?

We are starting a rollout to the US first but more to come! 🙏

great updates!

Hey @addyosmani one thing that occurred to me is will we be able to tell for analytics purposes when these agentic tools are interacting with webpages?

@opera is still winning the AI browser market. That stock might 10x this year lol. Super long $OPRA

It’s really good. It’s ripping me away from safari lol

Hopefully helps speed up a few tasks!

Oof. Google is on a murderous rampage

can’t wait to test these out f impressive

Hope they're helpful! 🙏

What's really cool is the Google Apps integrations, like Gmail. With memory and Personal Intelligence it will truly be essential to people deep into Google ecosystem. Great job!

The sidebar AI wars are heating up

Very cool

🙏🙏

@googlechrome Sounds interesting 🤓

Can it control the browser?

That's exactly what the auto browse feature we just announced does :)

Nice.

🙏

uhm..... it says no...

Sir, I don't want to be rude 🙏. But this whole thing looks silly. First make a UI for humans (using html,css, js), then use agentic browser to navigate the UI. The earlier UI (html,css,js and abstractions built over it) was built for humans (visual, motor)....ctd

Can't wait for my agentic bowser to doomscroll on x.

Auto-browse is the real story here. Not "AI in the browser" — but "AI that IS the browser." Think about it: every browser extension, every SaaS dashboard, every repetitive web workflow — all replaceable by an agent that can see and act on web pages. The browser just became an OS for AI agents. Chrome isn't adding features. It's becoming the platform.

It looks better than I expected. I wonder what the security concerns would be here.