At its annual I/O developer conference, Google unveiled Beam, a next-generation 3D video communication platform that builds on its earlier Project Starline. Beam uses artificial intelligence to convert traditional 2D video feeds into immersive 3D experiences, enabling more lifelike and intuitive virtual conversations.
The platform integrates seamlessly with existing workflows using Google Cloud, aiming to offer enterprise-level performance. According to the company, Beam could soon support real-time speech translation, enhancing cross-language communication. Initial Beam-enabled devices will launch later this year in partnership with HP, targeting enterprise customers.
How Google Beam Works
Beam captures users from multiple angles using a multi-camera setup and processes the video streams using AI to generate a 3D light field display. This setup, combined with precise head tracking (accurate to the millimeter at 60fps), allows users to maintain eye contact, perceive depth, and interpret subtle visual cues.
Google’s AI volumetric video model ensures that 2D streams are rendered with a strong sense of depth, delivering realistic 3D visuals that adjust to the viewer’s perspective. This creates a virtual communication experience that feels closer to in-person meetings.
Originally introduced in 2021 as Project Starline, the technology aimed to enable life-size 3D video calls with spatial audio and eye contact. Though the original project didn’t go mainstream, its core innovations have been refined and reintroduced through Beam.
Google has also confirmed that real-time translation features will be integrated into both Beam and Google Meet starting today, further enhancing accessibility and global collaboration.
The first OEM-produced Beam devices will be officially showcased at InfoComm 2025, scheduled for June.