Google I/O 2025: All AI products announced last night, from Beam to Veo 3

AI dominated Google I/O 2025 as the company launched Beam, Imagen 4, Veo 3 and real-time translation across Gemini, Android, Gmail and Meet

At its annual developer event, Google I/O 2025, the tech giant unveiled a range of new products and features powered by artificial intelligence. From lifelike 3D video calls with Beam to advanced image and video generation tools like Imagen 4 and Veo 3, Google showcased how AI is shaping the future of communication, creativity, and everyday tech. The company also introduced Android XR for smart wearables and new updates to Gemini and Search, aiming to make technology more helpful, personal and immersive.

 

Google unveils Beam: 3D video conferencing set to redefine virtual meetings

 

At its I/O 2025 developer conference, Google officially introduced Beam, a next-generation 3D video conferencing system designed to replicate the feeling of in-person interactions. Formerly known as Project Starline, Beam has moved beyond the research stage and is now being prepped for commercial use in workplaces.

 

 

Beam uses a combination of AI, advanced 3D imaging and specialised displays to render life-size, high-fidelity images of participants—creating a sense of physical presence without the need for headsets or glasses. The technology is designed to preserve natural eye contact, gestures and non-verbal cues, offering a more immersive and human communication experience.

 

The system captures standard video feeds and converts them into 3D representations that can be viewed from multiple angles. This enables users to better perceive subtle facial expressions, voice nuances and body language—elements often lost in traditional video calls. Beam operates on Google Cloud and integrates with existing enterprise tools, easing deployment for businesses.

 

To bring Beam to the market, Google is partnering with HP, which will launch the first Beam devices later this year. These devices will debut at the upcoming InfoComm trade show. Compatibility with popular platforms like Google Meet and Zoom is also planned, allowing users to join Beam sessions without changing their preferred conferencing software. 

 

At #GoogleIO, we shared how decades of AI research have now become reality.  From a total reimagining of Search to Agent Mode, Veo 3 and more, Gemini season will be the most exciting era of AI yet.  Some highlights ???? pic.twitter.com/2n9rbGNj0Q

— Sundar Pichai (@sundarpichai) May 20, 2025

  Google is also collaborating with partners such as Zoom, Diversified and AVI-SPL to accelerate adoption. Early testers include major names like Deloitte, Salesforce, Citadel, NEC, Hackensack Meridian Health and Duolingo. Commenting on Beam’s potential, Angel Ayala, Managing Director at Deloitte Consulting, said it is “not just a technological breakthrough but a reimagining of how we connect.”

 

Additionally, Google is expanding its video communication features with real-time speech translation. This functionality, already rolling out to Google Meet and set to come to Beam, allows for cross-language conversations with live translations that maintain original voice tone and inflection—enhancing clarity and connection in global meetings.

 

With Beam, Google aims to make virtual communication feel more authentic, enhancing both how people appear and how they interact during calls. More updates and expanded availability are expected in the coming months.

 

Gemini Live rolls out globally with visual understanding features

 

Gemini Live, Google’s conversational AI with vision capabilities, is now available for free to all Android and iPhone users starting today. The feature incorporates advancements originally introduced under Project Astra at I/O 2024, which enabled the AI to interpret objects through a device’s camera and respond contextually to user queries.   

 

Google Search gets ‘AI Mode’ for more conversational, visual results

 

Starting this week, users in the US will begin to see a new AI Mode integrated into Google Search and Chrome. Described as part of a broader overhaul of how search works, the new mode allows for chatbot-style interactions, where users can ask follow-up questions and engage in natural language exchanges instead of sifting through a list of links.

 

AI Mode also includes visualisation tools that can process complex data and generate custom charts. Google plans to enhance personalisation by using past search history and integrating with services like Gmail to surface more relevant content within search results.

 

Android XR: Google’s new OS for smart wearables

 

Google introduced Android XR, a new operating system tailored for extended reality (XR) devices, including headsets and smart glasses. Touted as the “first Android platform built in the Gemini era,” Android XR is designed to support immersive, AI-powered experiences.

 

Drawing from Project Astra, Gemini on Android XR enables devices to share the user’s point of view and respond accordingly—whether answering questions, performing tasks or providing real-time assistance. During its demo, Google showed Android XR in action through smart glasses that could send messages, schedule events, offer live navigation and translate languages instantly. One striking feature was the appearance of live subtitles during multilingual conversations.     

 

Google launches Imagen 4 and Veo 3 for image and video generation

 

At I/O 2025, Google debuted Imagen 4, its most advanced image generation model to date. Available immediately in the Gemini app, Imagen 4 features improved rendering of text and visuals and shows marked improvements in spelling and typography accuracy.

 

Alongside this, the company also released Veo 3, the latest version of its video generation model. Veo 3 can now produce not just realistic visuals from text prompts, but also accompanying audio such as ambient sounds and character dialogue.

 

We’re also launching Veo 3, our state-of-the-art video generation model.  Veo 3 lets you generate videos with sound effects, background noises and even dialogue.#GoogleIO pic.twitter.com/atVx6HvI9R

— Google Gemini App (@GeminiApp) May 20, 2025

  “Veo 3 excels from text and image prompting to real-world physics and accurate lip syncing. It’s great at understanding; you can tell a short story in your prompt, and the model gives you back a clip that brings it to life,” Google said in a blog post. Veo 3 is available today for Ultra subscribers in the US via the Gemini app and Flow, as well as for enterprise users on Vertex AI.

 

AI personalisation coming to Google apps

 

Google is enhancing user experience by introducing more personalised AI across its app ecosystem, beginning with Gmail. The upcoming “Personalised Smart Replies” feature will allow users to draft emails that reflect their tone and context more accurately.

 

This capability will be available later this year to paying customers and is part of Google’s broader initiative to tailor AI experiences based on user-specific context drawn from various Google services.