We finally have a real look at Google’s Android XR glasses in action

We’ve been tracking the development of Google’s Android XR platform for a little while now. Designed for the next generation of smart glasses and mixed-reality headsets, we got the first conceptual glimpse late last year directly from Google. While those more-produced videos showcasing potential use cases are interesting, they often paint an idealistic picture that is a bit removed from real-world performance.

However, things recently got a lot more tangible. During the TED2025 conference, Google’s Shahram Izadi presented the first live demonstration of Android XR running on a pair of prototype smart glasses, and thankfully, the video of that demo is now available for everyone to see. While we previously knew this demo was shown off at TED2025, there was no actual footage to see. Now there is.

Up until now, our main look at Android XR on glasses was that December video from Google. It showed potential, sure, but seeing it work live and on stage is a different beast entirely. In the TED demo, Izadi, assisted by Nishtha Bhatia, showcased the capabilities using prototype glasses featuring a display lens – which Izadi noted can support prescriptions. He mentioned using them for speaker notes on stage and confirmed they connect to your phone, leveraging all your phone apps.

The demo itself started simply enough, with Gemini generating a haiku on demand via the glasses. But things ramped up quickly. In a truly impressive display reminiscent of the Project Astra capabilities now rolling into Gemini Live, Nishtha turned briefly away from a bookshelf, and several seconds later asked the glasses-integrated Gemini if it caught “the title of the white book on the shelf behind me.” Gemini recited it instantly. This same real-world awareness was then used to locate a misplaced hotel key card on the shelf. The demonstration continued to showcase practical, AI-driven features:

• Visual Understanding: Explaining the contents of a diagram shown to the glasses.
• Real-time Translation: Translating a sign into English, and then subsequently into Farsi (Persian).
• Seamless Multilingual Interaction: Nishtha then spoke to Gemini in Hindi without changing any settings, and Gemini replied instantly and accurately in Hindi. This seamless language switching is particularly noteworthy.
• Contextual Actions: Looking at an album, Gemini identified it and offered to play a song from it.
• Navigation: A heads-up display showed navigation directions overlaid with a 3D map.

These examples move beyond simple notifications and hint at a genuinely helpful, context-aware AI assistant integrated directly into your field of view. While the glasses shown are clearly prototypes, this live demo provides the most concrete look yet at Google’s vision for ambient computing through eyewear. Seeing these AI features all working in a live environment – especially the real-time awareness and translation – adds a layer of credibility and excitement that can’t be matched by a concept video.