
Google has just given Gemini hands. Starting today, November 19, 2026, the Gemini app for Mac can do more than answer questions: it can operate your computer. A new feature called computer use lets Gemini carry out multi-step tasks across your local files, desktop apps and the web — clicking, typing, opening menus and filling in forms on your behalf, while you keep working on something else entirely.
The announcement, published on Google’s Keyword blog by Michael Friedman, Group Product Manager for the Gemini App, frames it as a time saver for repetitive work. “Just describe what you need, and Gemini handles the clicking, typing, menus, and forms for you,” Google says. It is the latest step in Gemini’s shift from a chatbot you talk to into an agent that does things for you — a shift that has accelerated sharply this month.
What Gemini Computer Use on Mac Actually Does
Computer use is not a new chat mode. It is a way of pointing Gemini at a job and letting it drive the software itself. You describe the outcome you want in plain language, and Gemini works through the steps: opening the right app, navigating its menus, typing into fields, moving files around and pulling information from websites.
To start a task, you type @Computer use in your prompt and describe what you want done. From there, Gemini takes over the mechanics. The important distinction is that it is doing the work inside real applications, not just telling you how to do it. That is the same idea behind so-called computer-use agents from other AI labs, but this is the first time Google has shipped it inside the Gemini app on a desktop.
The feature arrives just two days after Google made its other Mac agent, Tasks, free — a sign of how quickly the company is stacking agentic features into the desktop app. The Gemini app for Mac itself is only a few months old: Google released the native macOS app on April 15, 2026.
How the Background Window Works
The cleverest part of the design is that Gemini does not take over your screen. It works inside an isolated background window, with a live preview of what it is doing shown right inside your Gemini chat. Your main screen and your cursor stay completely free, so you can keep typing, designing or taking calls without interruption.
That separation matters more than it sounds. Earlier generations of computer-use agents grabbed your mouse and keyboard, which meant you had to sit and watch — or walk away — while the agent worked. Google’s approach turns the agent into a colleague working in the next room: you can check in on the live preview whenever you like, and your own work never stops.

What You Can Use It For Today
Google’s announcement lists four kinds of work to try first, and they are deliberately chosen to span different apps:
- Speed up creative workflows: batch color-grade photos in your favourite design software, or set up complex 3D scenes.
- Skip manual form entry: extract weights, dimensions and addresses from local PDFs to fill out multi-box shipping quotes.
- Turn live data into ready-to-present decks: pull metrics and screenshots from live dashboards across multiple regions and build a presentation in Google Slides.
- Batch-list items for sale: analyse clothing photos from a local folder, research pricing, log item details and stage listings on resale platforms.
The common thread is work that crosses app boundaries — the kind of chore where the information lives in one place, the tool lives in another, and a human spends most of their time copying between the two. That is exactly where an agent that can operate the computer itself earns its keep.

Who Can Use It — and What It Costs
Computer use is rolling out in English in the U.S. in the Gemini app for Mac, and it is not free. Access requires a Google AI Plus, Pro or Ultra subscription, with plans starting at $4.99 a month for AI Plus. It is limited to personal Google accounts for users aged 18 and older.
On the hardware side, you need an Apple Silicon Mac running macOS Sequoia 15.0 or later. Intel Macs are out, and Google has said nothing about a Windows version — so far, computer use has been announced only for the Mac app. The Gemini app for Mac is available to download at gemini.google/mac. As with most Google rollouts, the feature is arriving gradually, so not every eligible account will see it on day one.
The Controls Google Built In
Handing an AI the keys to your computer is a bigger trust decision than asking it to write an email, and Google has built in several controls to manage that. You choose which apps Gemini can access. You can watch every step live in the preview, pause the work at any time, and approve key actions before they happen.
Those controls are not optional extras. Google’s own documentation cautions that Gemini can make mistakes — including permanently deleting files or sending and sharing them with third parties — so the approval step and the ability to stop the agent mid-task are doing real work. When Gemini works with computer files, it creates temporary backups you can recover before moving to another task, though those backups are deleted when a new query starts or after 24 hours.
The timing is also notable: the launch comes as Apple prepares tighter rules for Mac apps that request full disk access, according to The Verge — a reminder that the whole category of agents that can operate your computer is about to face much closer scrutiny from platform owners.
How It Fits the Bigger Agent Race
Gemini’s computer use does not arrive in a vacuum. Google’s own Tasks agent, which went free earlier this week, already handles web-based errands like signing into sites, making purchases and booking appointments. OpenAI’s computer-use agents are now working inside hospitals, clicking through electronic health records and insurer portals. And on the hardware side, Meta’s Muse Charm puts a personal AI agent on a keychain, ready to talk without a phone at all.
What makes Google’s version distinct is the combination: a desktop agent that works in the background, inside the apps people already use, without hijacking the screen. Google’s Gemini 4 Neon model is also rolling out this month, giving Gemini a faster near-frontier engine to work with. The open-source Cowork project shows the same idea spreading in a different direction — a shared task queue that lets Claude Code, Codex and Gemini CLI hand work to each other. The industry is converging on the same conclusion: the next phase of AI is not a better chatbot, it is software that operates other software.
What to Watch Before You Hand Over Your Mouse
A few caveats are worth keeping in mind. Google’s announcement gives no benchmarks, success rates or task-completion times, so there is no public evidence yet for how reliably computer use finishes long tasks without human rescue. Google also does not say which Gemini model powers the feature — the announcement names no model at all. And the rollout is limited: U.S. English only for now, paid subscribers only, Apple Silicon only, with no date given for wider availability.
That is normal for a first release, but it means early adopters should treat computer use as a supervised assistant rather than an autopilot. Start with low-stakes tasks — the photo batch, the form fill — and watch the live preview. Save the multi-app workflow that touches your finances or your files for after you have seen how it behaves.
Even with those limits, the direction of travel is clear. In the space of a week, Gemini on the Mac has gone from answering questions to running errands to operating the computer itself. The mouse, it turns out, was never the hard part. Trust was — and Google is betting that a background window, a live preview and an approval step are enough to earn it.
Sources
- Google — Automate Mac tasks with Gemini computer use: https://blog.google/innovation-and-ai/products/gemini-app/computer-use-mac/
- TechCrunch — Google gives Gemini the ability to use your Mac for you: https://techcrunch.com/2026/11/19/google-gives-gemini-the-ability-to-use-your-mac-for-you/
- MacRumors — Google’s Gemini Can Now Control Apps on Your Mac: https://www.macrumors.com/2026/11/19/google-gemini-mac-app-control/

