Google unveiled a new browser-based AI experiment on Thursday called "Disco," powered by Gemini. This tool transforms your open browser tabs into customizable applications. With Disco, users can create what Google calls “GenTabs”—interactive web apps that proactively suggest ways to enhance your browsing tasks and allow you to build personalized tools using simple text prompts.
For instance, if you're studying a specific topic, GenTabs might recommend creating a visual web app to help clarify and reinforce key concepts.
In less academic scenarios, GenTabs could assist in generating a meal plan from a collection of online recipes or help organize an itinerary while you research travel options.
While similar outcomes are possible today with certain AI chatbots, GenTabs leverages Gemini 3 to instantly construct these tailored experiences by combining real-time browser data with your Gemini chat history. Once built, these apps can be further refined using natural language commands.
Google emphasizes that all elements generated through the GenTabs experience will include links back to their original sources.
Like other players in the AI space, Google is working to embed AI more deeply into the web browsing experience. Rather than launching a standalone AI browser—such as Perplexity’s Comet or ChatGPT Atlas—it has opted to integrate its Gemini assistant directly into Chrome, enabling users to ask questions about the webpage they’re currently viewing.
GenTabs goes beyond just the current page, focusing instead on broader, multi-tab browsing workflows—whether for research, learning, or other complex tasks.
Initially, this feature will be available only to a limited group of testers via Google Labs, who will provide feedback to shape its development. Google notes that compelling ideas emerging from Disco may eventually be incorporated into larger Google products.
This also signals that GenTabs is just the first of many planned features for Disco, marking it as the initial capability currently under evaluation.
To try Disco, users must join a waitlist to download the app, which is launching first on macOS.
