Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

MacRumors

macrumors bot
Original poster
Apr 12, 2001
70,752
42,652


AI-powered browsing features have become a battleground among browser makers in recent months. OpenAI launched its Atlas browser in October, while Microsoft Edge and Opera have also introduced AI assistants and automated browsing capabilities. And now Google is upping the ante. The company has announced a slew of new AI features for Chrome, including a persistent sidebar for the Gemini chatbot and "auto browse" capabilities that can perform web tasks on your behalf.

Still_Gemini-in-Chrome-Side-Panel-1.jpeg

The update moves Gemini from its previous floating window into an always-available side panel on the right side of the browser, which inevitably shrinks the browsing window. Google, however, says the redesign allows you to multitask more easily, since you can keep your main tab open while using the AI assistant for separate tasks, such as comparing products across multiple tabs or summarizing reviews from different websites.

Chrome is also gaining Nano Banana integration, Google's AI image generator. You can now transform images directly in the browser window using text prompts, without needing to download files or switch tabs.

The headline feature though is "auto browse," which is currently rolling out to AI Pro and Ultra subscribers in the U.S. According to Google, the agentic tool can handle multi-step tasks like researching hotel and flight prices, filling out online forms, scheduling appointments, and managing subscriptions. Google says it can even identify items in a photo, search for similar products, add them to a shopping cart, and apply discount codes. The company says it even does all this while staying within a specified budget.

Auto browse can use Chrome's built-in password manager (with user permission) for tasks requiring login credentials. The feature is designed to pause and ask you for confirmation for sensitive actions like purchases or social media posts, according to the company.

chrome-nano-banana.jpeg

Lastly, Google plans to add "Personal Intelligence" to Chrome in the coming months. The feature lets the browser remember context from past conversations you've had with it and provide more personalized assistance.

Google hasn't specified a Chrome version that will include all of these features, but the company said the Gemini sidebar support and Nano Banana integration are rolling out now as a server-side service update.


Article Link: Google Chrome Gets Gemini Side Panel and Agentic Browsing Features
 
Last edited:
I have been testing the paid version of Gemini, and while the overall offering seems quite reasonable (a lot of AI tools plus 200 GB of cloud storage), I don’t really understand the hype around the Gemini LLM itself. I don’t see major differences compared to ChatGPT in general use, whereas over the past week roughly every fifth query ended with an endless loading spinner, as if the model had frozen or Google’s infrastructure couldn’t keep up with generating responses. Maybe Gemini is slightly more versatile than ChatGPT, but it lacks a desktop app for macOS, which ChatGPT has and which, in my opinion, works great - always accessible via a keyboard shortcut, with automatic integration into VSCode or Terminal. This speeds up my work far more effectively than Gemini’s better benchmark results.
 
  • Like
Reactions: 0bit
those features are incredible, frankly. the relutance towards AI/llm needs to be mitigated and more selective. not everything is useless and stupid
Choice is what is needed. The choice to turn it on or off as one pleases. The current AI on google search results cannot be turned off without going through many technical hoops, and neither was I offered a choice to opt-in in the first place. I hope that this chrome feature offers that option, and to remain as the user chose, on or off, upon chrome restarting
 
Remember when all of this crap was a seperate "add-on" and "plugin" for your browser?

God, I'm sick of Google's constant assault and forcing Gemini upon people like a creepy incel demanding a date over and over and not being able to take 'no' for an answer.
 
Man, how hard is it for people to execute basic tasks like hotel room bookings themselves rather than relying on unreliable agentic task execution? Not once while doing something like hotel or flight research or online shopping have I craved automated control of that workflow. My own curiosity leads me to dive deeper into, say, certain hotel features or flight timings or whatever item I’m perusing, and I am chronically short for time and still don’t find said tasks to be fatiguing in any manner.

Glad I’m a safari user! Hopefully when and if Apple adds something like this, it’s in a Siri overlay and not a forced plugin.
 
  • Like
Reactions: Admiral Fart
Technically great but this is not the new OG Google moment at all. AI assistants need an intuitive overhaul and better UIs.
 
Man, how hard is it for people to execute basic tasks like hotel room bookings themselves rather than relying on unreliable agentic task execution? Not once while doing something like hotel or flight research or online shopping have I craved automated control of that workflow. My own curiosity leads me to dive deeper into, say, certain hotel features or flight timings or whatever item I’m perusing, and I am chronically short for time and still don’t find said tasks to be fatiguing in any manner.

Glad I’m a safari user! Hopefully when and if Apple adds something like this, it’s in a Siri overlay and not a forced plugin.
No kidding. I just booked a hotel yesterday, in the exact neighborhood I wanted, close to the exact restaurants I wanted, at the exact price I wanted, with the right amenities I wanted. It took all of 15 minutes, and no AI plug-in, no matter how advanced and intrusive, would have been able to get the exact result I wanted.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.