If you’ve ever wanted to have a normal conversation with Google — not the clipped “Hey Google, weather” type, but a back-and-forth chat that listens, looks, and responds in real time — that future is here. Google has begun rolling out Search Live, a multimodal, voice-first way to use Search that’s available now to English-language users in the U.S. through the Google app and Lens.

Search Live lets you tap a new Live button under the search box in the Google app (Android and iOS) and talk to Search as if you were chatting with a human assistant. It will answer out loud, show a running transcript, surface relevant web links, and — crucially — you can turn on your camera so Search can “see” what you’re looking at and follow up in context. Think: point your phone at a pile of cables and ask “Which goes to the monitor?” or show the tea whisk and scoops you bought and ask “Which of these is the right tool for matcha?”

How to try it

Open the Google app and look for the Live icon beneath the search bar. Tap it, start talking, and if you want visual help, hit the “Video” control to share your camera feed. The conversation stays open, so you can ask follow-ups without retyping, and you’ll get on-screen links if you want to dig deeper. Search Live is also reachable from the Lens app, giving the same back-and-forth camera + voice experience.

Where this came from

Search Live wasn’t invented overnight — Google had been testing conversational, multimodal search inside Google Labs and in earlier experiments in AI Mode. That opt-in period let Google tune the real-time UI and safety controls before widening the rollout. Now the company says the wider U.S. launch removes the Labs opt-in and opens the feature to (English) users more broadly.

Google’s broader “AI Mode” work runs on its Gemini family of models, and reporting around the launch ties Search Live into Gemini-style live conversation tools (Gemini Live). In short, the responses you hear are generated by Google’s generative models, with web links added to let you verify and explore the sources behind the AI’s answers. That mix — generative AI answers plus web citations — is central to how Google hopes to keep the experience useful and grounded.

Search Live converts what used to be a handful of typed keywords into a continuous, multimodal conversation. That’s huge for quick, hands-free tasks (cooking, troubleshooting, comparing products in a store), for accessibility (voice + visual context helps users who can’t type), and for the general shift away from search pages toward assistant-style interactions. It also raises questions about how result attribution, commerce links, and SEO will evolve — if the first answer comes from an AI voice readout, how will publishers compete for attention? Early signs are that Google will still surface web links alongside spoken answers, but the form factor is different and attention is scarcer.

Google itself points to low-friction, everyday uses: hands-free trip planning, step-by-step help for hobbies and schoolwork, and troubleshooting electronics by pointing the camera at ports and cables. Reviewers and testers note that the tool is handy when you don’t want to fumble with the keyboard — you can ask follow-ups naturally and keep the context intact.

A few practical limits matter right now: the rollout is targeted at U.S. users set to English, and some features that rely on device capabilities may vary by phone. More importantly, you’re opting into a live microphone and (optional) camera feed: Google’s support pages explain how Search Live stores a transcript and how you can review AI Mode history, and they document camera/mic controls and when the feed is used to produce an answer. If you’re privacy-minded, check the permissions and the AI Mode history settings — the experience is convenient, but it’s not the same as a private local search.

Google says Search Live will expand over time (and some reporting suggests India and other markets are next). Expect ongoing tweaks to make the AI more accurate, safer, and better at surfacing trustworthy links — and expect publishers, advertisers, and browser makers to pay close attention as the product shapes how people look for answers by voice and camera.