Model Releases

LongCat-Video-Avatar-1.5: upgraded audio-driven digital human video generation framework

Original title:meituan-longcat/LongCat-Video-Avatar-1.5

AI summary

Meituan's LongCat team released LongCat-Video-Avatar-1.5, an open-source framework focused on audio-driven digital human video generation. Its core upgrade adopts the Whisper-Large audio encoder, significantly improving the smoothness and naturalness of lip dynamics. The version achieves precise lip sync, full-body temporal stability, and identity consistency in long videos, and generalizes to complex scenarios such as anime, animals, and multi-person interaction. Through DMD2-based step distillation, the model requires only 8 steps for efficient inference. The team also built a human evaluation benchmark covering multiple scenarios and languages, validating its strong performance across key dimensions through large-scale subjective ratings and expert analysis.

Key Takeaways

  • Meituan's LongCat team released LongCat-Video-Avatar-1.
  • 5,an open-source framework focused on audio-driven digital human video generation.
  • Its core upgrade adopts the Whisper-Large audio encoder,significantly improving the smoothness and…
  • The version achieves precise lip sync,full-body temporal stability,and identity consistency in…
  • Through DMD2-based step distillation,the model requires only 8 steps for efficient inference.

Quick Facts

Published
Updated
Source
美团 LongCat:HuggingFace 新模型
Category
Model Releases
AI score
73
Read the original Read the full article at the original source
Why it matters

美团把数字人模型升级到1.5版,换了Whisper做音频编码,唇形同步比之前自然不少,而且开源了训练代码,做电商直播和虚拟博主的朋友可以直接拿过来跟商业方案掰手腕。

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: LongCat-Video-Avatar-1.5: upgraded audio-driven digital human video generation framework. May 21, 2026. https://new.nat6.net/en/news/cmpfoox6408bdsljwnbdhvasf/

View all →