Google’s disco browser signals a shift toward AI-driven web interaction.
Google has introduced Disco, an experimental AI-powered web browser designed to convert natural-language prompts and browsing context into task-specific web applications. Built on the company’s latest Gemini 3 model, Disco explores a new approach to web interaction by embedding generative AI directly into the browsing experience.
Unlike conventional browsers that rely on manual navigation across multiple tabs and websites, Disco aims to streamline complex online tasks through a feature called GenTabs. This capability analyses a user’s open tabs, prior searches, and written prompts to generate interactive, purpose-driven mini applications within the browser. For instance, a user planning a trip can prompt Disco to compile an itinerary that integrates maps, schedules, and relevant links into a single, organised interface.
The objective, according to Google, is to reduce the cognitive overhead associated with managing fragmented information across multiple web pages. Disco is built on the Chromium platform, allowing it to retain standard browser functionality while introducing AI-driven features at the interface level. Interaction with the system occurs through a chat-style sidebar, where users can initiate, refine, and adjust generated applications using natural language.
GenTabs remain connected to their original web sources, enabling users to trace information back to its origin and continue standard browsing when needed. Google has positioned Disco as a research-focused project under Google Labs, emphasising experimentation rather than immediate commercial deployment.
Disco is currently available in limited early testing via a waitlist for macOS users. While Google has not confirmed plans for a wider release, elements of the project, particularly GenTabs, could inform future enhancements across the company’s broader product ecosystem.