Alibaba Cloud launches Wan3.0 video model with document input and 30-second clips

Alibaba Cloud announced the public beta of its new video generation model Wan3.0 on the evening of August 6. The model upgrades in generation length, versatile creation, comprehensive reference, and realism, allowing single clips of up to 30 seconds with complex camera movements like continuous panning and one-shot takes, helping users move from generating a single shot to telling a complete story. Wan3.0 is the first to support document formats like doc, xls, ppt, pdf, and md, letting users upload materials such as lecture slides, product demos, dynamic charts, and generate videos with prompts, making it easy to turn office content into videos. The model also offers smart duration suggestions and video extension features, automatically adapting to prompts and extending the story. For human portraits, Wan3.0 focuses on details to show natural expressions and movements, avoiding the greasy AI look. In comprehensive reference tasks, it maintains consistency in characters, props, scenes, and styles. Starting today, Wan3.0 is available for public beta on Alibaba Cloud Bailian, Wanjing Yike, Wanxiang official site, and Qianwen Creation PC, with Qianwen app in gray release. API pricing is 0.3, 0.6, and 1.2 yuan per second for 480P, 720P, and 1080P, with full API access coming soon.