Base44 Superagents-AI agent that does it all
Mini toolsLongCat Video Avatar
By MeiGen AI
Model family: LongCat
LongCat-Video-Avatar is an expressive audio-driven human video model from Meituan’s LongCat team. It produces long-duration, lip-synced avatar videos with consistent identity and rich, natural body motion, supporting text-audio-to-video, audio-text-image-to-video and video continuation modes, including multi-person setups. It introduces Disentangled Unconditional Guidance, Reference Skip Attention and Cross-Chunk Latent Stitching to avoid static poses in silence, copy-paste artifacts and error accumulation in long sequences.
New
Video
Gen 4
Overview
LongCat-Video-Avatar is Meituan’s audio-driven avatar model built on LongCat-Video, generating super-realistic, lip-synced long videos from audio plus optional text and images, with stable identity, natural motion and support for multi-person scenes.
Tools using LongCat Video Avatar
No tools found for this model yet.
Last updated: February 25, 2026
