Gemini on Mac gets computer use to run tasks in your apps and sites
Google added computer use to the Gemini app for Mac on November 19, 2026. Gemini can now click, type, go through menus and fill in forms in Mac apps and websites to finish multi-step tasks, working in a background window while you keep using your screen. It is rolling out in English in the US to Google AI Plus, Pro and Ultra subscribers, which brings Google’s desktop app in line with the agent features rivals already ship.
What’s new
In Google’s announcement, Michael Friedman, a group product manager on the Gemini app, writes that “Gemini can now navigate Mac apps and websites in the background to finish complex tasks.” You describe the workflow in the chat and Gemini does the clicking and typing across local files, desktop apps and the web.
- Runs in the background. The work happens in an isolated window, with a live preview inside the chat, so your own screen and cursor stay free.
- You stay in control. You choose which apps Gemini may access, can watch every step and pause at any time, and approve key actions before they happen. TechCrunch adds that you can also stop a task.
- Example jobs from Google: color-grading a batch of photos or setting up 3D scenes in design software; reading weights, dimensions and addresses from local PDFs to fill in multi-box shipping quotes; pulling metrics and screenshots from dashboards into a Google Slides deck; and turning a folder of clothing photos into staged resale listings after researching prices.
Who can use it
| Requirement | Detail |
|---|---|
| Mac | Apple silicon, macOS Sequoia 15.0 or later |
| Account | Personal Google account, 18 or older |
| Plan | Google AI Plus, Pro or Ultra |
| Region and language | US only, English only, rolling out |
The feature is part of the Gemini app for Mac, downloaded from gemini.google/mac. Google hasn’t said when it will reach other countries, languages or platforms, according to TechCrunch.
What it means for developers
Google’s post is about the consumer app. It doesn’t name the model behind the agent and doesn’t mention an API or any way for developers to build on the Mac agent. For now this is something to use rather than build on.
For developers who use it, the controls matter most. Gemini only reaches the apps you allow, so keep terminals, password managers and anything with production credentials off that list until you have seen how it handles tasks. The approval step before key actions is your last check before something like a form submission, so read it rather than clicking through. The examples Google picked, moving data from files into forms, dashboards into slides and batches of files through the same steps, suit the kind of repetitive internal work that never got its own script.
How it compares
TechCrunch notes that Google is late here: OpenAI, Anthropic and Meta already offer agent features in their desktop apps. The Mac launch follows two days after Google made Gemini Tasks free, the agent that can make calls and purchases for you on the web. Computer use takes that same idea to native Mac apps and local files, but only on paid plans.