Apple added Live Recognition to the iPhone and Vision Pro last year, and roughly a year later, Google has its answer.
Guided Vision launched in Gemini Live today, giving Android users real-time audio descriptions of their surroundings. The feature relies on your smartphone’s camera with no extra hardware required.
How does Guided Vision actually work?
To enable Guided Vision, turn it on in the Gemini app’s profile toggle, set an Accessibility Shortcut (like a two-finger swipe or both volume keys), or use a three-finger tap in TalkBack.
Share your camera during a Gemini Live session, and it narrates what it sees. So, reading fine print like nutrition labels or washing machine dials, or finding a dropped earbud, Guided Vision can do it all.
To help you place the camera correctly, Gemini talks you through reframing it. It tells you to pan right, tilt down, or step back to fit something in the frame. Beyond general descriptions, you can also ask follow-up questions, such as having it read an expiration date after spotting one.
To ensure the accessibility tool works across multiple languages, Google has tested it in multiple languages and regions, including India, Brazil, Singapore, and Japan.
Who built this, and what won’t it do?
Google trained Guided Vision with Aira, logging tens of thousands of hours of visual interpretation and leaning on over 1,000 Aira Trusted Testers to stress-test the model, while Aira specialists helped shape its safety guardrails.
It explicitly isn’t a medical device, mobility aid, or cane replacement, and Google says not to rely on it for navigation or obstacle detection, guidance that matters since a wrong description could put someone at risk.
It’s rolling out now on Android 9 and up, arriving as Google’s answer to the Live Recognition feature Apple shipped on iPhone and Vision Pro. For those catching up, Google previewed Guided Vision earlier this month alongside other Android accessibility tools.
RELATED COVERAGE:
- Google Gemini now lets you turn repeated workflows into reusable Skills
- Google Chrome now lets you take interactive quizzes and keeps your work synced across devices
- Google Pixel 11 can now make phone calls for you with Gemini
- Adobe just made Claude far more useful for PDFs by bringing Acrobat tools directly into the chat
