Skip to content
Apixo
Blog
news· 3 min read· via TechCrunch AI

Google Brings Computer Use Capabilities to Gemini Desktop Application on macOS

Google has launched computer use features for its Gemini macOS app, enabling the AI to control native apps, browse the web, and execute multi-step tasks in the background.

Google Brings Computer Use Capabilities to Gemini Desktop Application on macOS

Google announced on Thursday that it is adding computer use capabilities to its Gemini desktop application on macOS. The update allows the artificial intelligence to perform tasks directly on behalf of users, interacting with native Mac applications and navigating websites in the background.

With this release, Google joins a growing segment of the industry focused on agentic desktop tools. Competing model providers such as Anthropic, OpenAI, and Meta have already introduced desktop agent capabilities, positioning Google's release as an effort to close the gap with existing alternatives.

How Gemini operates on macOS

The new computer use functionality allows Gemini to carry out multi-step workflows spanning local files, desktop software, and web browsers. Users supply instructions through the chat interface, after which Gemini begins interacting with the system.

Rather than taking over the physical mouse pointer or locking the display, Gemini runs inside a live preview window embedded within the chat interface. This setup leaves the user's screen and cursor free, allowing them to continue working uninterrupted while the AI executes its instructions. Users also retain operational control through built-in safeguards: the interface provides options to pause, halt, or approve Gemini's actions prior to execution.

According to Google, the capability is aimed at streamlining repetitive and tedious computer work. Stated use cases include filling out online forms, applying repetitive actions across batches of files, sorting files into directories based on designated rules, and assembling slideshows using information sourced from local documents or online dashboards.

Google's expanding agentic roadmap

The introduction of computer use arrives alongside other recent agentic rollouts from Google. Earlier in the week, the company initiated a broader distribution of Gemini's "Tasks" feature, an in-app tool designed to act as an agent by assisting users with logging into websites, scheduling appointments, and completing online purchases.

In addition, Google recently opened access for early testers to evaluate its redesigned "CC" AI agent. First announced in September, CC focuses on family organization by managing information across calendars, emails, task lists, and messaging threads. While CC and Tasks concentrate on coordinated personal management and web actions, computer use extends Gemini's scope directly into the local operating system environment.

What it means for developers

For developers, Google's addition of computer use to macOS marks a significant step toward making operating systems programmable through natural language agents. As frontier model providers build tools capable of interacting with standard user interfaces—clicking buttons, reading interface elements, and navigating between separate software environments—the way users and automated tools interact with desktop software is evolving.

Instead of relying solely on dedicated APIs or complex custom scripting for automation, multi-step workflows can increasingly be delegated to model-driven agents capable of parsing visual and UI hierarchies. At the same time, this paradigm places higher demands on developers to evaluate which underlying foundation models offer the best reasoning and task execution for agentic behaviors. For developers interested in testing and integrating different frontier systems, platforms like Apixo (https://apixoai.online) provide cheap pay-per-token API access to models from Google, Anthropic, OpenAI, and others through a single unified key.

Availability and system requirements

At launch, Gemini's computer use feature is available exclusively in English within the United States. To access the functionality, users must install the Gemini desktop application for Mac and maintain an active Google AI Plus, Pro, or Ultra subscription tier.

Google has not yet detailed a timeline for when computer use might be released in international regions or made available on operating systems outside of macOS.


Source: Google gives Gemini the ability to use your Mac for you — TechCrunch AI. Written by the Apixo team from that report.

#ai-news#google#gemini#macos#ai-agents#automation
Try it with your own tools

One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.

Get your API key

Keep reading