Google shows off its AR smart glasses

Google gave a live-onstage demo of its AR smart glasses. Google PM Nishtha Bhatia stood on stage, asked Gemini for directions to a coffee shop, and the glasses used AR to display a map to direct her to the coffee shop. Oh yea, NBA star Giannis Antetokounmpo was wearing them onstage too. Check out the full story here.

Android goes extended reality

Google gave an update on its Android XR platform, the version of Android built for headsets and other extended reality device form factors. 

The company said hundreds of developers are building for the XR platform, which Google released in preview last year. Google’s Gemini AI assistant is a part of XR, giving context in apps and more. And Google said that its partnership with Samsung to build an XR headset, Project Moohan, will bear fruit later this year. Moohan will be available for purchase sometime in 2025.

Google’s stock is down slightly

Investors seem unimpressed by Google’s announcements during I/O 2025. Alphabet’s stock was down 1.2% during the keynote. However, other tech stocks, including Meta and Nvidia, were also down on Tuesday morning.

Audience groans for Gemini’s Ultra pricetag

At Google I/O 2025, the audience let out an audible groan when the $250-per-month price tag for Gemini Ultra flashed across the screen. That’s a lot of money for an AI subscription, even compared to OpenAI and Anthropic’s priciest plans. We mentioned this earlier, but in case you missed it, check out the full story on Gemini Ultra here.

Flow is Google’s new app to create AI films

Google debuted Flow at I/O – a new app with a collection of generative AI tools for making video clips. 

Flow incorporates a range of the search giant’s AI technologies, including its Veo 3 video-generating model. Users can take advantage of camera controls, storyboarding features, and more to create compelling, cohesive (more or less) short films. Check out the full story here.

Google says Synth ID has been used to watermark over 10 billion pieces of content

Two years ago, Google created Synth ID, which embeds invisible watermarks into AI-generated media. At Google I/O, the company said over 10 billion pieces of content have now been watermarked with the technology. Google is also expanding its partnerships to ensure that more content is watermarked with Synth ID and more organizations can detect the mark. 

At the event, Google said it’s making it easier to detect the watermark. The company’s new Synth ID detector can now identify if an image, audio track, text, or video has Synth ID in it – whether it’s in the whole piece or even just a part. This will begin to roll out to early testers today.

Google expands access to its Lyria music-generating models

Google is making its Lyria 2 AI model, which can generate high-fidelity music and audio, available to developers, enterprises, and more.

Lyria 2 is now in YouTube Shorts and Google’s dev-focused Vertex AI platform. Alongside it, Google has rolled out Lyria RealTime, a model that lets users interactively create, control, and perform music in real time. Lyria RealTime is in the Gemini API as well as Google’s AI Studio suite.

Google is upgrading its most capable Gemini AI models

Google announced an “enhanced” reasoning model called DeepThink at the Google I/O 2025 conference. Demis Hassabis, head of Google DeepMind, Google’s AI R&D org, told TechCrunch that  “[Deep Think] pushes model performance to its limits.” Check out the full story here.

Google Play is getting audio previews and other improvements

Remember when Google made a number of Android-related announcements at a pre-show before the start of its annual developer conference, Google I/O? You might recall that Google did hold back a bit on any news about the Play Store

Today, Google finally shared that it is rolling out a series of upgrades to its Play Store to help Android app developers better market their software and services to consumers. Check out Sarah Perez’s story here.

Gemini is now in Chrome

Google has brought its Gemini AI assistant to Chrome. Josh Woodward, VP of Google’s Labs group, said that Gemini can “understand the context” of web pages that you’re on automatically, so if you have a question, it can be answered.

“This will be your AI assistant that’s there for you as you browse the web on your desktop,” he said. “I especially love it for comparing reviews on long pages, like [a] camping website.”

Gemini in Chrome will be available this week for Gemini subscribers in the U.S.

Canvas uses Gemini to help you vibe code

Woodward also introduced Canvas, which he described as a Gemini-powered interactive space for co-creation. In a demo, he showed how Canvas lets him transform a detailed, written report into all kinds of new things, including a dynamic web page, an infographic, a helpful quiz, and even a custom podcast in 45 languages. 

Plus, he said Canvas can be used to vibe code all sorts of things, too, with as much back and forth as you want. 

After you create an app using vibe coding techniques, you can share it with others who can easily “jump in and view it and modify it and remix it,” he said. “This is the power to transform anything, and it’s a whole new way to use Gemini.”

Google has added lots of extras to its new video-generating AI model: Veo 3

Google unveiled a new and improved video-generating AI model called Veo 3. The improvements, which you can read about in more detail here, are pretty incredible. The company claims it can generate sound effects, background noises, and even dialogue to accompany the videos it creates.