New Delhi: Google’s DeepMind has recently released an Gemini 2.5 Computer Use model with the specialized version of its Gemini 2.5 Pro AI that can actually interact with the users interfaces the model is also available in the preview through the Gemini API through the Google AI Studio and Vertex AI Studio, and these model are pretty much advanced and Google has provided an impressive interface for the users. This model enables the AI agents to complete tasks by interacting directly with graphical interfaces, such as filling forms, clicking buttons, scrolling and operating behind logins.
The ability to natively fill out forms, manipulate interactive elements like the dropdowns and the filters, and operate behind logins is the crucial next step in building powerful, general-purpose agents. Developers’ access to the model through the computer use tool, which operates in the loop. Inputs include the user request, a screenshot of the environment and a history of recent actions. The model generates responses in the form of UI actions, which are executed by the client-side code. The loop continues with the updated screenshots and context until the text is completed or terminated.
The model is optimised for web browsers and shows potential for mobile UI control, but is not yet designed for desktop operating system-level tasks. Demonstrations include transferring pet care data to a CRM system and organising digital sticky notes into categories. Gemini 2.5 Computer Use has demonstrated strong performance on web and mobile control benchmarks, including the Online Mind2Web, WebVoyager, and AndroidWorld. According to some sources, the model is going to deliver high accuracy while maintaining low latency, with the accuracy above 70% and latency around 225 seconds. Google DeepMind emphasised safety, noting that AI agents controlling computers carry risks such as misuse, unexpected behaviour and web-based scams.
