Alibaba's Wan3.0 Video AI Doubles Generation Length To 30 Seconds
Wan3.0 turns office documents into video sources, letting users generate a full product PV or narrated report from slides and spreadsheets without assembling multiple short clips.
Reporting from 1 source: GIGAZINE.
Alibaba began general availability of the Wan3.0 video generation AI on August 24. The model reads text, images, documents, spreadsheets, slides, and web pages, and creates videos up to 30 seconds in a single generation, double the 15-second limit of the previous Wan2.7-Video. It includes a video extension function and carries over the editing tools introduced in Wan2.7.
The 30-second ceiling matters because most AI video models stop at a few seconds per generation. Product introductions and narrative ads usually require several clips stitched together, and matching faces, clothing, backgrounds, and props across those cuts becomes the hard part. Wan3.0 aims to remove that assembly step by generating a full scene series, movement, dialogue, and camera work, in one pass.
The document reading is the other change. Supported inputs include doc, xls, ppt, pdf, txt, key, pages, numbers, and md, plus web page links, one file or link per request up to 100MB and 50 pages. Alibaba pitches turning product slides into ad videos and spreadsheets into animated graphs.
Wan3.0 is available from Alibaba Cloud Model Studio and Qwen Cloud. The Standard version is discounted 30% until September 23, 2026.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.