Introducing the Gemini 2.5 Computer Use model, a significant advancement for developers leveraging the Gemini API. This specialized solution enhances interaction with user interfaces (UIs) by harnessing the robust visual understanding and reasoning capabilities of Gemini 2.5 Pro. Outperforming competitors in web and mobile benchmarks, it boasts lower latency, making it ideal for digital tasks such as filling and submitting forms. Agents can now seamlessly navigate applications like humans—clicking, typing, and scrolling—while operating behind logins and manipulating interactive elements like dropdowns and filters. This functionality is crucial for creating versatile, general-purpose AI agents. With the new computer_use tool, developers can input user requests, screenshots, and action histories, tailoring UI actions as needed. Access these transformative capabilities through Google AI Studio and Vertex AI, and elevate your applications with cutting-edge AI technology.
