Gemini API's Computer Use tool is now generally available for browser, desktop and mobile agents

On October 30, 2026, Google's Gemini API changelog announced that the Computer Use tool is generally available across the Interactions and GenerateContent APIs. The tool lets a model look at screenshots and return actions such as clicks, typing, scrolling, drag-and-drop and key presses, which your own code then carries out.

Google's docs list Gemini 4 Argon, Gemini 3.8 Flash (recommended for computer use), Gemini 3.5 Flash-Lite and Gemini 3 Flash Preview as supported models, across browser, desktop and mobile environments. Mobile adds actions such as opening and listing apps and long presses. Computer Use is not available on the free tier; on the paid tier it is billed as regular tokens at the chosen model's standard rate.

Each proposed action can carry a safety decision: allowed, blocked, or requiring confirmation. Built-in policy categories cover financial transactions, account creation, legal terms and data changes, and when confirmation is required your app must ask the user and send an acknowledgement back. Opt-in prompt-injection detection, off by default, scans screenshots for hidden instructions on Gemini 3.5 Flash-Lite and later.