Between 2023 and 2025, Vivo has actively expanded into the AI digital human field alongside its smartphone and MR hardware businesses. The company launched an AI Digital Human app that allows users to easily generate short videos with digital human avatars, and has recruited engineers focused on virtual digital humans. Vivo’s research teams, including its AI Global Research Institute and OS division, have invested heavily in large language models and generative AI tools to support digital human creation, live streaming, and real-time interaction. Applications range from sign language digital human translation systems (which won first place in national evaluations), to executive avatars in e-commerce livestreams, and to personalized digital human assistants integrated with smartphones, MR headsets, and future 6G-connected environments. Corporate white papers, such as Digital Life 2030+, position digital humans as part of long-term scenarios for education, entertainment, healthcare, and the metaverse.
The Vivo AI Global Research Institute, founded in 2018 with branches in Shenzhen, Beijing, Hangzhou, Nanjing, and San Diego, is dedicated to advancing research in computer vision, natural language processing, knowledge representation, and multimodal AI to support Vivo’s long-term integration of artificial intelligence across devices and services. Led by Vice President Zhou Wei, the institute has developed the Lanxin Large Model Matrix, including the Lanxin 3B end-side model optimized for mobile with high efficiency and low power use, alongside upgraded cloud models for speech, image, and multimodal tasks. These technologies power applications like PhoneGPT and are embedded in OriginOS 5, while its sister Communications Research Institute complements this work with 6G research, publishing multiple white papers and prototypes exploring AI-native networks and immersive services.