Skip to yearly menu bar Skip to main content


MAGVIT: Masked Generative Video Transformer

Lijun Yu · Yong Cheng · Kihyuk Sohn · José Lezama · Han Zhang · Huiwen Chang · Alexander G. Hauptmann · Ming-Hsuan Yang · Yuan Hao · Irfan Essa · Lu Jiang

West Building Exhibit Halls ABC 213
award Highlight
[ ] [ Project Page ]


We introduce the MAsked Generative VIdeo Transformer, MAGVIT, to tackle various video synthesis tasks with a single model. We introduce a 3D tokenizer to quantize a video into spatial-temporal visual tokens and propose an embedding method for masked video token modeling to facilitate multi-task learning. We conduct extensive experiments to demonstrate the quality, efficiency, and flexibility of MAGVIT. Our experiments show that (i) MAGVIT performs favorably against state-of-the-art approaches and establishes the best-published FVD on three video generation benchmarks, including the challenging Kinetics-600. (ii) MAGVIT outperforms existing methods in inference time by two orders of magnitude against diffusion models and by 60x against autoregressive models. (iii) A single MAGVIT model supports ten diverse generation tasks and generalizes across videos from different visual domains. The source code and trained models will be released to the public at

Chat is not available.