Gemini Live's new Guided Vision describes the world out loud for blind and low-vision users

Point your Android phone's camera and Gemini Live now describes what it sees, reads fine print and tells you to tilt or step back when the shot is off. Google built it with Aira and more than 1,000 testers.

By ScoviAI Team

Published | Updated · 2 min read

Google Guided Vision in Gemini Live announcement image
Google Guided Vision in Gemini Live announcement image (Google)· 2 min read

Key points

  • Launched in early October 2026 on compatible Android devices
  • Share your camera in Gemini Live to get spoken descriptions and reframing cues such as 'pan slowly right'
  • Trained with Aira on tens of thousands of hours of visual interpretation; 1,000+ testers helped refine it
  • Works in multiple languages; turn it on in the Gemini app or via an Android accessibility shortcut

For people who are blind or have low vision, a camera app that "describes the scene" is only useful if the camera is pointed at the right thing. Google's new Guided Vision tries to fix that part too.

The feature is now available in Gemini Live on compatible Android devices. You share your camera, and Gemini talks you through your surroundings like a patient friend on a video call.

How Guided Vision works

Turn it on, open Gemini Live and tap to share your camera. From there you can ask questions out loud and get spoken answers. The new part is what Google calls reframing cues: if the camera is too high, too close or off to one side, Gemini tells you to pan slowly to the right, tilt down or step back until it can see clearly.

Google gives several everyday examples:

  • Reading fine print: nutrition labels, washing machine dials, printed menus in a dim restaurant.
  • Finding things: a dropped earbud on the floor, the black pepper in a crowded spice cabinet.
  • Matching details: checking whether a striped shirt goes with a pair of trousers.
  • Getting the layout: a quick overview of an unfamiliar room or what's spread across a table.

Built with the people who will use it

Google developed Guided Vision with Aira, a service that connects blind and low-vision people with trained visual interpreters. Together they interpreted tens of thousands of hours of visual data to train the model. More than 1,000 members of Aira's Trusted Tester network stress-tested it in daily routines, and Aira specialists helped set and evaluate the safety guardrails.

The feature works across several languages, and Google says it was tested with blind and low-vision communities in India, Brazil, Singapore, Indonesia and Japan, among others.

How to turn it on

There are two ways:

  1. In the Gemini app: open your profile settings and switch on Use Guided Vision in Live, then start Gemini Live and share your camera.
  2. As a shortcut: in Android Settings, go to Accessibility, then Vision assistance, then Guided Vision, and set up an accessibility shortcut.

Why it matters

Visual-description apps have existed for years, but most take a single photo and describe it. A live, two-way conversation that also coaches you on aiming the camera is closer to having a sighted helper. Google notes it can help older adults, people with low literacy or anyone reading small text in poor light. As with any AI description, it can get things wrong, so it's best treated as a helper, not as the final word on things like medication labels.

Google's newest model behind many Gemini features is covered in Gemini 4 Argon is now available.

Sources

Source: Google

ScoviAI Team is the ScoviAI newsroom. We write every article ourselves from the original announcements and link our sources. Spotted a mistake? Tell us and we'll fix it.

Related Topics

ADVERTISEMENT