Fri 20 Nov 2026 / 02:33 ET
Kernel
AI 3 min read

Google rolls out Gemini computer use for Mac in the US

Google says Gemini can handle clicks, typing and forms on eligible Macs, though the action-taking feature has a limited US rollout.

Theo Lindgren

By Theo Lindgren / Columnist

Google rolls out Gemini computer use for Mac in the US
img: The Verge

Google on November 19 began rolling out Gemini computer use Mac, an action-taking feature that the company says can help users finish multi-step work involving local files, desktop applications and websites. It is a separate addition to Google’s Gemini app for macOS, and it comes with narrower eligibility than the basic Mac app.

According to Google’s announcement, a user describes a task in Gemini, after which the app can handle interface work including clicking, typing, opening menus and entering information in forms. Google presented that as a way to reduce repetitive sequences that might otherwise take many manual steps. The company has not supplied independent evidence in the available material about how reliably the feature completes those jobs.

What can Gemini computer use on a Mac do?

Google says the feature can work across approved local files, Mac apps and the web. Its examples include extracting weights, dimensions and addresses from PDFs to populate shipping-quote forms; assembling metrics and screenshots from dashboards into a Google Slides presentation; preparing resale listings from clothing photos; and handling parts of photo-editing or 3D-design workflows.

Those are Google’s proposed uses, not verified performance claims. The company says Gemini carries out the work in an isolated background window, with a live view of its activity displayed in the chat. Google says this leaves the user’s primary screen and cursor available for other work.

What controls do Mac users have?

Google says users choose which applications Gemini may access. The company also says users can watch the live activity, halt it at any point and must approve key actions before they occur. Those are the stated controls for this rollout; the announcement does not provide a detailed list of every action that needs approval or every action the software may be barred from taking.

The permission model matters because the feature is intended to interact with data and interfaces outside a chat window. Google’s description establishes that app access is selected by the user, but it does not independently establish the feature’s security or accuracy in real-world work.

Who can use Gemini computer use on Mac?

The capability is rolling out in English for users in the United States, Google says. It requires a personal Google account, a user aged 18 or older, an Apple Silicon Mac running macOS Sequoia 15.0 or later, and a Google AI Plus, Pro or Ultra subscription.

That should not be confused with the broader Gemini macOS application. Google previously made the native app available globally at no cost for Gemini users on macOS 15 and later. That app can be summoned with an Option-Space shortcut and can accept shared screen or window context, including local files, to answer questions about what a user is viewing. Google’s desktop page also lists voice dictation across open apps and organizing local files into Google Docs and Sheets among the Mac app’s capabilities.

Computer use goes beyond supplying context to Gemini: Google says it is designed to let the service operate interface controls after the user grants access. For now, the practical catch is the rollout’s US, English, hardware, age and subscription gates.

This story draws on original reporting from The Verge.

More AI/

view all ↗