Alibaba's Wan3.0 adds 30-second video, audio, and document inputs at $6 per 1080p clip
AI video model releases
로그인하고 알림을 만들어 보시기 바랍니다.
Alibaba's Wan3.0 adds 30-second video, audio, and document inputs at $6 per 1080p clip
Alibaba Cloud's Wan3.0 is API-only and generates up to 30 seconds of video in one pass, versus 15 seconds for Wan2.7, with audio rendered in the same generation.
The Omni-Reference input system accepts one document up to 100MB and 50 pages, including PDF, Word, spreadsheets, slides, text files, Apple formats, or one public URL.
Wan3.0 accepts up to 10 reference images, five reference video clips totaling 15 seconds, and five reference audio clips totaling 15 seconds; video input and output share a 30-second total budget.
List pricing is $0.20 per second at 1080p, making a 30-second clip cost about $6; 480p and 720p cost $0.05 and $0.10 per second.
Alibaba published no model weights or local deployment path, and the article says no independent benchmark has been published.
cho.sh