One Model, Three Modalities: ByteDance Releases Lance for Image and Video Understanding, Generation, and Editing

2026-05-20 22:14 GMT · 2 months ago aimagpro.com

ByteDance’s Intelligent Creation Lab has released Lance, an open-source native unified multimodal model that handles image and video understanding, generation, and editing — all within a single framework, using only 3B activated parameters.
The post One Model, Three Modalities: ByteDance Releases Lance for Image and Video Understanding, Generation, and Editing appeared first on MarkTechPost.