NVIDIA has released its XR AI developer library in public beta, giving developers an open framework to build multimodal AI agents that operate on augmented reality glasses, XR headsets, and extended reality devices. The announcement was made at the Augmented World Expo (AWE) 2026 in Long Beach, California — the premier annual gathering for spatial computing — where NVIDIA’s David Chu delivered a keynote showcasing the platform’s transformative capabilities.
The XR AI platform bridges the physical and digital worlds by connecting lightweight XR wearables to an organization’s GPU resources across cloud, data center, workstation, and edge deployments. According to NVIDIA’s developer documentation, the library enables AI agents that can “find, guide, verify, and capture in real-world complex enterprise environments” — capabilities that open entirely new possibilities for hands-free AI assistance in industrial, medical, and field service contexts.
The first commercial hardware built on the platform was unveiled at AWE by VITURE: the Helix AI safety glasses. Designed for industrial, scientific, and clinical environments, Helix combines a transparent safety-glasses form factor (meeting ANSI Z87.1-2025 standards) with a 12-megapixel camera, four-microphone array, stereo speakers, Wi-Fi, Bluetooth 5.3, and over 60 minutes of battery life. The device streams the wearer’s first-person perspective to a multimodal AI in real time, enabling coaching, compliance monitoring, and documentation capture without requiring the worker to stop their task.
VITURE Helix is expected to begin shipping to enterprise customers in Q1 2027, with pricing starting at $600. Enterprise pilot allocations opened on an invite-only basis at AWE. This early pricing suggests that AI-powered safety eyewear is approaching commercial viability at a scale that could make it standard equipment in manufacturing, healthcare, and field service operations within the next few years.
AWE 2026 runs June 15–18 under the theme “I, Spatial,” spotlighting the convergence of spatial computing and AI that NVIDIA’s platform enables. The public beta release builds on NVIDIA’s previous collaborations with VITURE and Stanford’s Cong Lab, demonstrated at GTC 2026 in March, where the partners showcased AI-powered lab automation workflows using XR glasses and robotic arms — providing a glimpse of how AI and spatial computing will reshape scientific research.
The open nature of the XR AI developer library is particularly significant. By releasing it as an open framework, NVIDIA is positioning its GPU infrastructure as the backbone for an entire ecosystem of XR AI applications — from industrial safety to medical training to retail assistance — without locking developers into a proprietary stack. This approach mirrors the strategy that made CUDA ubiquitous in AI development and could similarly establish NVIDIA’s XR AI platform as the standard for the next generation of wearable AI.
