Google adds Computer Use to Gemini Mac app to run desktop tasks in the background

Google now lets Gemini handle Mac tasks in the background. The new "Computer Use" feature in the Gemini Mac app completes multistep tasks across local files, desktop apps and websites, according to The Decoder's report on the announcement.
Google says users simply describe a goal, and Gemini clicks, types, and navigates menus and forms. Everything runs in an isolated background window with a live chat preview, which keeps the screen and cursor free so the user can carry on working.
The user stays in control. Users choose which apps Gemini can access, watch every step live, pause at any time, and approve key actions beforehand.
Google's examples include batch-editing photos and 3D scenes, filling out shipping forms from local PDFs, and building presentations.
Computer Use rolls out in English in the US on the day of the report, and the Gemini Mac app is available to download. The Decoder adds that OpenAI has also focused on Computer Use recently, and says it is pulling well ahead with its Astra model.
Key facts
- Google added a "Computer Use" feature to the Gemini Mac app that completes multistep tasks across local files, desktop apps and websites.
- Users describe a goal; Gemini clicks, types and navigates menus and forms, all inside an isolated background window with a live chat preview.
- Users pick which apps Gemini can access, watch every step live, pause anytime and approve key actions beforehand.
- Google's examples: batch-editing photos and 3D scenes, filling out shipping forms from local PDFs, building presentations.
- Rollout is in English in the US; The Decoder says OpenAI is pulling well ahead in Computer Use with its Astra model.
Why it matters
This moves Gemini from answering questions to operating a Mac on the user's behalf, across local files, desktop apps and websites. The isolated background window is the notable design choice: the agent works on its own while the screen and cursor stay free. The Decoder also notes that OpenAI has focused on Computer Use recently and is pulling well ahead with its Astra model, so Google is entering a contested area.
Who it affects
Mac users who run the Gemini Mac app and want routine desktop chores handled in the background. Google's examples point at people who edit photos or 3D scenes in batches, fill out forms from local PDFs, or build presentations. For now the rollout covers English in the US only.
How to use it
Download the Gemini Mac app and describe a goal in plain language. Choose which apps Gemini can access, follow the steps in the live chat preview, pause when needed, and approve key actions beforehand. Computer Use rolls out in English in the US. The source states no pricing, subscription tier or system requirements for Computer Use or the Gemini Mac app.
How solid is it
The account rests on a short report from The Decoder describing Google's announcement. The description of what Gemini does, "clicks, types, and navigates menus and forms", is attributed to Google. The source gives no benchmark results, success rates or limitations for the feature, and it does not name the model behind Computer Use. The Decoder's claim that OpenAI is pulling well ahead with Astra is the outlet's own statement, with no comparison data given.
Risks and caveats
An agent that clicks and types in local apps raises the stakes of mistakes, which is why the controls matter: app selection, live viewing, pause and approval of key actions. The source describes no security or privacy mechanisms beyond those. The rollout is limited to English in the US, and no dates or plans for other languages or countries are given. The source makes no statement on Windows or other platforms.
“Everything runs in an isolated background window with a live chat preview, keeping the screen and cursor free for uninterrupted work.”
— The Decoder, describing Google's Computer Use feature