Computer Use Agents

Google

Gemini Computer Use

Computer use model on the Gemini API and Vertex AI, covering browser, mobile and desktop interfaces.

Google ships computer use as a tool in the Gemini API, documented across browser, mobile and desktop environments — mobile is the one no other entry here documents as a first-class target. Gemini 3.8 Flash is the recommended model as of October 2026. It has been in preview for a year, through four model generations.

Best for

Teams already on Google Cloud that want one model across web and native interfaces.

Watch out

Still preview, and Google's own documentation says so plainly: it may contain errors and security vulnerabilities, and it should not be pointed at critical decisions, sensitive data, or actions that cannot be corrected. A year of preview is a long time; do not read the model churn as the stage moving.

Changelog

  1. Gemini computer use reaches desktop and mobile, and stays in preview

Also in the directory

Computer use tool OpenAI's computer use tool on the Responses API, for developers building their own control loop. The earlier computer-use-preview path is deprecated. Copilot Studio computer use Computer use tool inside Copilot Studio. Automates Windows desktop applications and websites for agents, billed per step. Dots Persistent agents inside ChatGPT, each with its own cloud computer and browser. They keep working after you close the tab, and you can watch the screen while they do.