On November 19 Google added computer use to the Gemini app for Mac. Describe a task and Gemini clicks, types and works through menus and forms in an isolated background window to finish multi-step jobs across local files, desktop apps and the web, with a live preview in chat while your own screen and cursor stay free. Users choose which apps Gemini may access, can pause at any time and approve key actions. It is rolling out in English in the U.S. to Google AI Plus, Pro and Ultra subscribers and requires an Apple Silicon Mac on macOS Sequoia 15 or later.

Key Takeaways

  • ✓Official: Google blog post on November 19 by Michael Friedman, Group Product Manager for the Gemini app, plus a demo post from @GeminiApp the same day
  • ✓How it works: runs in an isolated background window with a live preview in chat, so you can keep using your own screen and cursor
  • ✓Control: users pick which apps Gemini can access, can pause at any time, and approve key actions before they happen
  • ✓Official examples: batch color-grading photos, setting up 3D scenes (the demo turns a sketch into a rigged Blender model), filling shipping forms from local PDFs, and building Google Slides decks from dashboards
  • ✓Availability: U.S. English, Google AI Plus, Pro and Ultra, personal accounts 18+, Apple Silicon Mac on macOS Sequoia 15.0+
Gemini App for Mac Gets Computer Use: Works Across Local Files, Desktop Apps and the Web in an Isolated Background Window
🖼️Official Media
Click to view high-res
🧭

Turn your technical choice into a development budget

Compare 40 dev plans & simulate token costs vs $20/mo subscriptions

🔬

In-Depth Technical Analysis

Background

A lot of daily work means clicking, copying and filling forms across local files, desktop software and websites, and chat assistants could only advise while a person still did the clicking. Computer-use agents are moving from developer APIs into consumer desktop apps, and Google is now putting this capability into the Gemini app for Mac.

How it works

You describe what you need and Gemini works through menus, types information and clicks through apps and websites to finish a multi-step task. It runs in an isolated background window with a live preview in the chat, so your main screen and cursor are left alone. For safety, you choose which apps Gemini can access, can watch every step and pause at any time, and approve key actions before they happen.

Benchmarks

Google published no benchmark scores or success rates. Its blog lists examples: batch color-grading photos in design software, setting up complex 3D scenes, extracting weights, dimensions and addresses from local PDFs to fill out shipping quotes, building a Google Slides deck from metrics on live dashboards, and analyzing clothing photos to stage resale listings.

What developers should do

Download the Mac app at gemini.google/mac. It is rolling out in English in the U.S. only, for personal Google accounts 18+ on Google AI Plus, Pro or Ultra, and needs an Apple Silicon Mac on macOS Sequoia 15.0 or later. Be careful about granting access to code editors or terminal-style apps: allow only the apps a task needs, and keep in mind that web content can carry prompt-injection risk.

Action HubReady to adopt this in production?

Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.

Up NextX Debate: AI Commentator Chubby Says Claude Code and Opus 5.5 Have Been His Daily Driver for Two Months and GPT Is "Not Close," but Concedes Computer Use Is Still GPT's Edge; 150+ Replies Focus on Splitting Work by Task and the Same-Day Claude Login Outage
Read Now→