• Source:JND
Add as a preferred

Google has taken another big leap in AI with the launch of Gemini 2.5 Computer Use, a new model designed to interact with the web much like a human would. Built on Gemini 2.5 Pro, this version isn’t just about generating answers—it’s about actually doing things online.

In a detailed blog post, Google explained that Gemini 2.5 Computer Use is “capable of interacting with user interfaces”, meaning it can perform real actions like clicking, typing, scrolling, filling out forms, and even opening dropdown menus. The company says this AI outperforms competitors across several benchmarks while also offering lower latency than rival models.

ALSO READ: Realme 15 Pro 5G Game Of Thrones Edition Is Out Now: Check Price In India, Specs And Features! Is It Worth It?

Developers can already access it through Google AI Studio and Vertex AI, where they can test and integrate its capabilities into apps or automation workflows. For now, the model supports 13 distinct actions, though Google notes that it can only operate inside a browser and “is not yet optimised for desktop OS-level control”.

To show how it works, Google released demo videos (sped up three times) where Gemini 2.5 Computer Use completes full tasks from scratch. In one example, it was prompted with:

“My art club brainstormed tasks ahead of our fair. The board is chaotic and I need your help organizing the tasks into some categories I created. Go to stick-note-jam.web.app and ensure notes are clearly in the right section. Drag them there if not.”

The AI navigates the webpage, drags notes into the correct categories, and finishes the task — exactly as a person would.

While still in its early stages, the potential here is massive. Google says its internal teams are already using the model for UI testing, which could make software testing significantly faster. The company also revealed that variations of this AI are powering agentic features in tools like AI Mode in Search, Firebase Testing Agent, and Project Mariner — an experimental platform where users can assign AI agents to handle everyday tasks such as research, planning, and data entry.

ALSO READ: Xiaomi 17 Pro Max, 17 Pro, And 17 Secure Top 3 Spots On AnTuTu V11 Benchmark

With Gemini 2.5 Computer Use, Google is moving closer to a future where AI doesn’t just understand what you want—it can actually go online and do it for you.