ByteDance targets July launch of upgraded AI video model
ByteDance Ltd.’s cloud computing unit announced Tuesday it will launch an upgraded artificial intelligence video generation model in early July, aiming to capture the professional filmmaking and commercial advertising markets. Volcengine’s Seedance 2.5, currently in late-stage beta testing, introduces native 4K resolution, 30-second video outputs, and sophisticated 3D pre-visualization tools. The release follows the February launch of Seedance 2.0, which rapidly became the dominant production engine for China’s booming micro-drama industry. Seedance 2.5 addresses the unpredictability of current AI video tools by giving creators precise control over spatial logistics. ByteDance CEO Liang Rubo said the model’s 3D pre-visualization feature, inspired by professional film directors, allows production teams to plan camera movements and lighting on early-stage virtual sets before rendering final textures.
