Google has officially entered the growing race for AI-powered web browsers with a new experiment called Disco. While several AI browsers have already made waves with features like agent-style browsing and built-in assistants, Google’s approach is a little different: Disco isn’t being positioned as a full, mainstream browser replacement yet. Instead, it’s launching through Google Labs as an early-stage testbed for next-generation browsing ideas.
Right now, Disco is limited to users in the United States and requires joining a waitlist. Google is starting with macOS access first, with Windows and Linux support expected to arrive later.
Google describes Disco as a “discovery vehicle” for exploring what the future of the web could look like when AI is deeply integrated into everyday browsing. The goal is to develop AI-powered utilities that help people navigate, research, and get things done online more efficiently, without needing technical skills or complicated setups.
The biggest feature being introduced with Disco is GenTabs. This tool looks at your open browser tabs along with your chat history inside the browser, then uses that context to generate something useful on the spot—such as an interactive tool or even a custom mini web app tailored to what you’re currently working on. In other words, instead of manually piecing together information across multiple pages (or copying content into separate AI tools), GenTabs tries to turn your browsing session itself into a smart workspace.
Under the hood, Disco runs on Google’s latest AI model, Gemini 3. Google’s pitch is centered on simplicity: you don’t need to code, install special extensions, or build anything from scratch. You just describe what you want in natural language, then refine it through conversation until the browser generates a personalized tool that fits your task.
Google’s early demos focus on familiar scenarios many people use AI for today—building a Japan travel plan, organizing weekly meal prep, or mapping out a garden layout. These examples make it easy to understand what Disco is trying to do, even if they resemble the same types of AI demonstrations that have been circulating across the tech world for the past couple of years.
What makes Disco interesting isn’t necessarily the specific examples, but the direction behind them: Google appears to be testing how AI can transform browsing from passive scrolling into active problem-solving, where the browser can assemble context, create interactive workflows, and reduce the need to jump between apps.
That said, it’s still hard to say whether Disco has a long-term future as a standalone AI browser. Because it’s a Google Labs experiment, it’s likely designed to test features, gather feedback, and iterate quickly. If any of these ideas prove genuinely useful, they could easily become future features inside Chrome—especially since Chrome remains Google’s flagship platform for most users.
For US users who want early access, Disco is currently available through a waitlist, with macOS support rolling out first and broader platform availability expected later.
