ByteDance is preparing an AI model that generates spatial video in real time. The work targets uses in robotics and self-driving systems. It would put ByteDance against Meta and Alphabet in that area. Founder Zhang Yiming is personally overseeing the project. He has pulled together teams from across the company. He has also directed computing capacity and other AI resources to it. A launch could come as soon as next month. The timing is not settled and the plans may change. The account comes from people familiar with the matter, speaking anonymously. ByteDance has not confirmed the model publicly.