Google DeepMind has integrated computer use into Gemini 3.5 Flash, allowing developers to create agents that can interact across different platforms. This enhancement enables reliable custom agent development for tasks such as continuous software testing and professional application automation. The feature is now accessible through the Gemini API and Gemini Enterprise Agent Platform.
Computer use is a built-in tool in Gemini 3.5 Flash, delivering the best performance for agentic computer use tasks. Previously, computer use was available only as a standalone Gemini 2.5 model. With this integration, developers can use 3.5 Flash to build agents that can see, reason, and take action across browser, mobile, and desktop environments. This capability improves long-horizon and enterprise automation tasks.
Gemini 3.5 Flash with computer use can analyze the Gemini app and return a categorized list of features. It also audits its own documentation for accessibility issues. To mitigate prompt injection risks for agents operating in live environments, targeted adversarial training is used for computer use in Gemini 3.5 Flash.
Source: deepmind
