As an open-source avatar generator, LongCat1.5 creates lip-synced talking videos from one picture and audio for humans, cartoons and pets. Optimized encoder and sampling enable fast run on 8G VRAM, supporting multi-character dialogue and video continuation for bulk short-video production.
longcat video avatar 1.5
longcat video avatar 1.5
It generates dynamic avatar videos via audio-driven images with precise lip sync, supporting real people & anime and fast inference on low VRAM.