On August 24, the video generation model Wan3.0 was officially launched. The model has been comprehensively upgraded in four dimensions: generation duration, universal creation, comprehensive reference, and realistic world rendering. It can generate up to 30 seconds of native video at one time. For the first time, it supports input of document formats such as doc, xls, ppt, pdf, and md. It is committed to accurately reproducing the real physical world and lowering the threshold for high-quality video creation.

QQ20260824-112036.jpg

Since the public beta test started on August 6, Wan3.0 has been tested by many creators and enterprise users. During the beta testing phase, the model continued to iterate and optimize, constantly improving in instruction following, cross-shot consistency, audio quality, and video editing capabilities.

In the field of short films, industry teams regard Wan3.0 as a core tool for large-scale production. Wang Jian, CTO of Jingying Technology, stated that the model performs excellently in close-up shots, with precise facial details and high stability and usability of the final product; Si Yu, co-founder of Juri Lu, mentioned that relying on a mature production platform, the audiovisual quality has been comprehensively upgraded, providing a new solution for large-scale short film production.

For the general creator platforms, Wan3.0 is called the "cost-performance king" due to its stable human-like quality, accurate reference identification, 30-second continuous generation, and reliable cross-shot consistency. Platforms like LibTV, Pollo.ai, and Zhaomeng Cicero reported that the model significantly reduces repeated debugging costs, allowing ordinary creators to independently complete complete narrative short films.

The advertising and marketing industry also gave high praise to Wan3.0. Chen Long, CTO of Guangzhou Chopsticks Information Technology, pointed out that the model ensures the uniformity of product appearance, logo, and material, with natural and smooth voice-over, and the final product is close to the level of real advertising shooting; the team of San Dai Technology verified that the model can quickly integrate stylized reference images, adapting to various advertising creative needs.

In addition, Wan3.0 has also been applied in the fields of cultural tourism and music MV. FLYAI uses movie-level lighting and camera movement to vividly present landmark scenes around the world; Lu Xiaoxu, CEO of Xiaoxu Music, introduced that only a single reference image is needed to achieve automatic storyboarding and lip-sync singing, and the final product can be directly used as a music MV. Overseas creators also participated in the evaluation. Mehul Agarwal, co-founder of Koyal.ai, stated that Wan3.0 ranked first in multiple animation evaluations, with outstanding performance in high-dynamic animation and realistic visuals.

Currently, users can experience Wan3.0 through channels such as Alibaba Cloud BaiLian, Wanxiang official website, Wanjing Yike, Qwen AI full-end, Duiyou, and IF STUDIO. API pricing is divided into three tiers: 0.3 yuan/second for 480P, 0.6 yuan/second for 720P, and 1.2 yuan/second for 1080P. From August 24 to September 23, Alibaba Cloud BaiLian and Qwen AI platforms are offering a 70% discount on API services.