At this week’s Google I/O developer conference, we had a chance to get a quick taste of Google’s upcoming AI-powered glasses. Rather than the audio-only glasses that the company announced will begin shipping this fall, these glasses offer a combined audio and visual experience.
First announced at last year’s event, these Android XR glasses feature an in-lens display that overlays the real world to display useful information right in front of your eyes. This includes widgets that can display weather, walking directions, Uber pickup details, live translation, and more. It also includes widgets that you design yourself using AI.
The company says the glasses will also work with iOS and Android phones, both in audio-only format and in a future display version.
The eyewear with a display is intended to be the next step beyond the first generation of audio glasses, which will be released later this year. Developed in partnership with Warby Parker, Gentle Monster, and Samsung, the glasses combine Google’s technology with the brand’s design aesthetic.
The glass we tested, on the other hand, was still more of a prototype, but has been polished enough to be tested externally. The person who demoed the XR glasses explained that the prototype freed Google from having to worry about the cosmetic details associated with different styles and shapes, and instead allowed it to experiment more freely with display technology and focus on its impact on battery life. This means that these glasses will be significantly different from future versions of the glasses in terms of fit, shape, dimensions, and attention to detail. Rather, it’s more like you can experiment with the “innards” of your glasses while still maintaining a basic, comfortable frame.
The shipping version of the glasses is designed to detect when you put the glasses on or take them off your head, but the one we tried didn’t have this feature.
To activate Gemini, press the right side of the glasses frame for 2 seconds. You’ll hear a startup chime, letting you know Gemini is on and listening. In the demo version, when you start Gemini, the camera also starts at the same time, but in the shipping version, users will be able to set whether or not to turn on the camera when Gemini starts.
For our first test, we asked Gemini to play our favorite artists and played the music through the glasses. Unfortunately, the venue was too noisy to evaluate the sound quality. The music was set at maximum volume, so it was still relatively difficult to hear clearly and in detail. However, my first impression from this limited experience was that the glasses are no substitute for high-quality earphones, although they can be useful if you just want some music while you’re out, walking, hiking, or doing chores around the house. The benefit of not wearing earbuds is that it makes it easier to hear what someone is saying compared to the transparent mode experience on devices like Apple’s AirPods.
To turn off the music, tap once in the middle of the side of the frame, like tapping your temple.
In the second test, we pressed the photo capture button to take a photo of a person. Since the display was turned off, the photos were transferred to my phone and watch. (You’ll later be able to capture a video with a long press, but this option wasn’t available for testing on the prototype. For videos, you’ll see a thumbnail preview of the video instead of a photo.)
You can also simply ask Gemini to take a photo and perform some AI operations on the result without pressing the photo button. For example, you can say, “Take a photo and turn that person into an anime character.” The photo is sent to your phone, then to Gemini and Nano Banana servers, and returned in an edited version.
The Wi-Fi at the Google I/O venue was heavily loaded and took about 45 seconds to make a round trip.
When you enable the display, a simple home screen appears in your field of vision. The demo version came preloaded with several widgets showing the weather and a countdown to Google’s I/O event. If a specific app, like Google Maps or Translate, is your primary use case for the glasses, you can also embed the quick launcher into that specific app.
The prototype only had one display in the right eye, but the platform can support both single and dual displays, as well as audio-only glasses. The image itself was a little blurry, but I attributed this to my prescription contacts. This requires wearing a distance-optimized lens on one side and a near-optimized lens on the other. When I closed one eye, the images focused better, but the experience left me with some eye strain over my right eye almost immediately, and I’m not sure if the prescription was entirely to blame.
One of the best demos is the language translation experience on the glasses, powered by the Google Translate app on your phone. When one of the demonstrators spoke rapid-fire Spanish, the glasses automatically detected the language and showed English text on the display, and Gemini spoke English into our ears. We see travelers all over the world buying glasses just for this experience.
Note that translation also works with audio-only glasses, just without the text displayed on the glasses. Instead, you can see the transcription on your phone if you want, in addition to real-time audio feedback.
Another demo involved using glasses to navigate. Obviously I couldn’t go for a walk or leave the venue to test the accuracy, but I was able to get an idea of how it works. You can start your Google Maps experience by asking Gemini to navigate you to your destination. The destination can be as vague as “the nearest coffee shop.”
Gemini launches Google Maps on your phone, and after a short delay while the experience loads, you’ll see turn-by-turn directions on your glasses. When looking ahead, the following sequential information is displayed: However, if you need to get your bearings in space, look down at the ground and see the blue dots on the map. You can also rotate it in space by rotating left and right, similar to trying to point the blue dot in the right direction on your phone.
Then, when you look up again, you can continue walking without the map getting in your way.
This experience is tied to Google Maps on your phone, so saved destinations like “Home” and “Work” are already available.
They were also able to briefly use the glasses to identify different objects in their field of vision and ask questions about them. The glasses initially had a hard time identifying the replica Monet painting on the shelf in front of them because the prototype’s camera didn’t automatically enable itself and had to be turned on again from the app. Still, even after we got closer to note Monet’s signature in the lower left, it took a few more questions before Gemini said it “looks like Monet.”
Other tests were smoother, as the glasses quickly identified plants on the shelf and answered questions about various recipes in the book. Still, these are things you can do today with Google Lens (or other AI models integrated into chatbot apps), and I think it’s interesting to be able to do them without taking out your phone.
Google says it plans to share more about Android XR display glasses when it expands its Trusted Tester program later this year.
In the meantime, the company believes audio is sufficient for some users’ needs, but this is probably a smart way to take advantage of the fact that display glass isn’t ready, even though it competes with Meta and Snap on this front.
Like the display version, the Audio Glasses also have access to Google’s Gemini AI, allowing you to listen to audio privately through the glasses’ frame speakers. Similar to future display versions, you can listen to music through the glasses, press a button to take a photo, make a call, and tap the Phone app from the glasses.
The use of other third-party apps wasn’t among the items we demoed, but the glasses allow users to tell Gemini to do things like “take an ingredient from this recipe and add it to your shopping list.”
In another example that Google showed off during the event’s keynote, the glasses can allow the wearer to watch food being cooked on the stove and provide feedback about the meal, such as whether the meat is fully cooked yet.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
