EZXR (易现科技) is a China-based AR and spatial-computing company that originated from NetEase’s AI business unit and positions its digital-human work as “AR-native” characters that can be created, customized, and placed into real physical environments with high-precision spatial anchoring. In practice, EZXR markets low-cost, configurable virtual characters (2D and 3D, including cartoon-to-realistic styles) with editable appearance, clothing, motion, voice, and text, and it describes combining conversational-AI-style interaction with TTS/ASR and rendering workflows so the digital human can speak and respond in real time inside AR experiences. Its flagship “guide” framing is a digital-human-enabled AR tour and service assistant for venues such as cultural tourism sites, commercial spaces, and exhibition environments, where a character is presented as an on-site explainer or concierge rather than a purely on-screen avatar. Technically, EZXR ties these digital humans to its broader “device–cloud” AR stack (including AR glasses support, 6DoF tracking, and real-time mesh reconstruction for occlusion/collision) so characters remain registered to the user’s surroundings and can be delivered as part of integrated hardware-plus-platform deployments.