Alibaba Debuts Wan 3.0 Video Model That Can Build Clips From Documents and Images Instead of Text Prompts
Alibaba Cloud has opened public beta testing for Wan 3.0, the newest entry in its Wan video-generation family, adding the ability to turn documents, images, audio and webpages directly into video without requiring a written prompt. The release, which went live on August 6, 2026 across Alibaba Cloud Model Studio, the Qwen Cloud platform, Wanxiang's official site, Qianwen Creation PC and other channels, doubles the maximum single-pass clip length to 30 seconds, up from the 15-second ceiling of the prior Wan 2.7 model, according to Alibaba Cloud's own product post.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.