Google I/O: All you need to know about Beam, Google's 3D video calling tech

At the Google I/O keynote, Google unveiled Google Beam, a next-generation 3D video communication platform designed to make remote conversations feel as natural as in-person meetings. Building on the foundation of Project Starline, Beam leverages artificial intelligence to transform standard 2D video streams into immersive 3D experiences on supported devices. Google also announced that select Beam capabilities—like real-time speech translation—are coming to Google Meet.

What is Google Beam?

 

Google describes Beam as an AI-first video communication platform that enables people to connect, collaborate, and communicate in 3D, with life-sized realism. Drawing on Project Starline’s 3D video technology, Beam is designed to power a new generation of AI-enabled devices, offering what Google calls a “true-to-life human connection.”

 

By capturing depth and detail—such as facial expressions, eye contact, and subtle gestures—Beam aims to make virtual conversations more engaging and emotionally rich. Users will be able to pick up on cues like a smile or a glance, making interactions feel less artificial.

How Google Beam works

 

At the core of Google Beam is a proprietary AI volumetric video model, which enables fully 3D video rendering from multiple angles. The system uses a six-camera array to capture the speaker from different perspectives. Google’s AI then merges these streams in real time, generating a volumetric 3D reconstruction that adds depth to the image.

 

On the receiving end, supported devices use a custom light field display to present a lifelike, three-dimensional representation of the caller—giving the illusion that the person is sitting across from you, rather than on a screen.

 

Google said that the model runs on Google Cloud, which enables the feature to offer enterprise-grade reliability, and compatibility with existing conferencing workflows.

 

Availability

 

Google has partnered with HP to develop the first devices built specifically for Google Beam. These devices will be officially unveiled at InfoComm 2025 next month and are expected to be available to select customers later this year. Google is also working with channel partners such as Diversified and AVI-SPL to expand Beam’s reach across businesses and organisations globally.

In addition, Google announced that Beam-powered real-time speech translation is coming to Google Meet. The feature delivers low-latency translations that preserve tone and expression. It is now available in beta for Google AI Pro and Ultra subscribers, with support initially offered for English and Spanish.