The Meituan LongCat team officially released and open-sourced a SOTA-level virtual human video generation model—Longcat-Video-Avatar. The model is based on the LongCat-Video base, continuing the core design of “one model supports multi-tasking”. It natively supports core functions such as audio-text-to-video, audio-text-image-to-video, and video sequencing. At the same time, the underlying architecture has been fully upgraded to achieve breakthroughs in the three dimensions of motion realism, long video stability, and identity consistency.

The Meituan LongCat team officially released and open-sourced a SOTA-level virtual human video generation model—Longcat-Video-Avatar. The model is based on the LongCat-Video base, continuing the core design of “one model supports multi-tasking”. It natively supports core functions such as audio-text-to-video, audio-text-image-to-video, and video sequencing. At the same time, the underlying architecture has been fully upgraded to achieve breakthroughs in the three dimensions of motion realism, long video stability, and identity consistency.