Meituan releases open audio-driven video model

thione · x · 2026-08-25

Meituan released LongCat-Video-Avatar 1.5, an open model for audio-driven human video generation with faster inference speeds.

Original post →

More from Multimodal

Multimodal channel →