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.
More from Multimodal
- Alibaba's Wan 3.0 launches: praised as major leap with prompt tips — OdinLovis · 2026-08-25
- AI filmmaking workflow: Creating a trailer in 2 days for XPrize — lmoroney · 2026-08-25
- Director Paul Trillo creates visual art with Adobe Firefly — paultrillo · 2026-08-25
- MM H3 2-Pass Latent Upscale Workflow: Eliminating Face Blur — Majestic-Ad-1336 · 2026-08-25
- MiniMax H3 Max launches with fal tuning for enhanced prompt adherence and visual quality — gorkem · 2026-08-25
- Pika launches WAN 3.0 with 20 references and 30-second generations — Mr_AllenT · 2026-08-25