ByteDance Reportedly Developing Real-Time Spatial Video Generation AI Model, Launch Possible as Early as Next Month

September 7 news, according to Bloomberg, ByteDance founder Zhang Yiming is personally overseeing the development of a new generation real-time spatial video generation AI model, which is expected to be released as early as next month. The model, built on ByteDance's existing AI video generation model Seedance, allows users to create interactive virtual worlds for live streaming, short dramas, and games, placing users in a three-dimensional space with a latency of about 0.05 seconds and 20 frames per second. Zhang Yiming hopes to leverage this to bring Douyin creators into the emerging field of "world models," competing with AI leaders such as Li Feifei and Yann LeCun. The model aims to reduce VR hardware costs by shifting computation-intensive content generation to the cloud. ByteDance is redirecting substantial B2B operations toward video production and generation, having previously secured a $30 billion loan for AI capability expansion.