Google’s Gemini can now use your Mac for you. The company announced on Thursday that computer use is coming to the Gemini app for Mac, letting the AI click, type, navigate menus and fill out forms across your files, apps and websites to finish multi-step tasks while you get on with something else.
Just describe what you need done, and Gemini handles the rest. According to Google, the feature “helps you get multi-step tasks done across your local files, desktop apps, and the web,” saving you “the repetitive work that normally takes dozens of clicks.”
What Gemini can actually do on your Mac
This is not another chatbot that answers questions about your files. Computer use takes action inside your apps. Google’s examples start with creative work — batch color-grading photos or setting up complex 3D scenes — and move quickly into the boring stuff that eats real afternoons: pulling weights, dimensions and addresses out of local PDFs into shipping quotes, grabbing dashboard metrics into a Google Slides deck, or analysing photos of clothes in a folder, researching prices and staging resale listings.
To start a task, you add @Computer use to your prompt. New “Desktop” apps such as Blender and Autodesk Fusion now appear in the app’s connectors, alongside Git and Obsidian — aimed at people who live in heavy desktop software.
It works in the background — your cursor stays yours
The cleverest part of the design is what Gemini does not do: it does not take over your screen. The AI works inside an isolated background window, with a live preview of what it is doing right in your chat. Your screen and cursor stay free, so you can keep working while it runs. That matters because the first generation of computer-use agents earned a reputation for hijacking your mouse and locking you out of your own machine for minutes at a time.
There are guardrails. You choose which apps Gemini can access, watch every step live, and pause it at any time. Key actions need your approval before they happen, and Google says it pauses for logins and payments. You are handing an AI the keys to your apps — but you still hold the master switch.
Who can use it — and who can’t
Here is the catch, and it is a big one: computer use is not free. It is rolling out in English in the United States only, and only to subscribers on a Google AI Plus, Pro or Ultra plan. You also need the Gemini app for Mac and a personal Google account, and to be 18 or older. Google has not said when it will reach other countries, languages or Windows.
The timing is telling. According to TechCrunch, Google only this week made its Tasks agent available to everyday users free of charge — while putting a paywall around the one that touches your files and apps.
Why this matters
That split is the real story of this launch. Google is drawing a line between AI that talks and AI that does — and charging for the doing. Tasks will happily plan your errands for free; the moment you want Gemini to open your apps, move your files and fill in your forms, you need a subscription.
It is also a catch-up move. TechCrunch notes that Google is “a little late to the game,” with OpenAI, Anthropic and Meta already offering agentic features in their desktop apps. Google spent 2026 building the chatbot; its rivals spent it building agents that act.
Whether it succeeds comes down to trust. An AI that can click, type and move files is genuinely useful — batch-processing a hundred photos or a folder of invoices is work nobody wants to do by hand. But it is also an AI with its hands on your keyboard, and Google’s background-window design, with live previews and approval gates, is its answer to that fear.
Meanwhile in Apps & Tools: Microsoft’s New Office App Puts Word, Excel and PowerPoint in One Copilot App
Meanwhile in Apps & Tools: YouTube’s Custom Feeds Let You Build Your Own Algorithm With AI


