AI News Feed
Market watch
LG UltraFine Evo 32U990A-S Hits Record-Low $1,099 on Amazon, TechRadar ReportsNewegg Labor Day Deal: 32GB Kingston Fury Beast DDR4-3200 Kit Nearly Half Price at $179.99 with CodeGalaxy Z Fold 8 inner screen corners reported with excessive giveHuawei unveils Mate XT 2 tri-fold with privacy-screen edition as Apple prepares folding iPhoneWall Street Analysts Recommend Nvidia, Uber and Marvell for Long-Term GrowthLenovo, AMD and Huawei Put Local Large-Model AI at Center of New Devices
LG UltraFine Evo 32U990A-S Hits Record-Low $1,099 on Amazon, TechRadar ReportsNewegg Labor Day Deal: 32GB Kingston Fury Beast DDR4-3200 Kit Nearly Half Price at $179.99 with CodeGalaxy Z Fold 8 inner screen corners reported with excessive giveHuawei unveils Mate XT 2 tri-fold with privacy-screen edition as Apple prepares folding iPhoneWall Street Analysts Recommend Nvidia, Uber and Marvell for Long-Term GrowthLenovo, AMD and Huawei Put Local Large-Model AI at Center of New Devices
Products & Applications

Alibaba Launches Public Beta of Video Model Wan 3.0, Supporting Document-to-Video

Alibaba's Wan 3.0 video generation model entered public beta on August 6, featuring 30-second single-generation, multi-document input support, and enhanced realism.

AI News Feed · 2026-08-14 22:27

The extended generation duration is the most visible change. A single 30-second output allows for complex camera movements and one-shot sequences, moving AI video from generating a single shot to telling a complete story. Wan 3.0 also includes an intelligent duration feature that recommends appropriate lengths based on prompts.

Wan 3.0 has evolved from a video generator to an expression carrier for information. It can take text, images, audio, and video as references, and also read structured information from documents and PPTs. Combined with prompts, it can produce teaching courseware, product demos, and business reports. Supported formats cover doc, xls, ppt, pdf, md, with file or link limits of 100MB and 50 pages. For example, uploading a PPT about smart glasses with a prompt yields a complete promotional video.

The model's strong instruction-following and prompt engineering capabilities serve as the hub that transforms prompts into scheduling inputs and controlled expression.

Video generation models are shifting from content generation to productivity tools, and users demand realism as well as clarity. Wan 3.0 aims to accurately replicate and present real scenes or digital information. Generated portraits strive for individuality, with sharper facial and skin details, natural and restrained emotional expressions, and coordinated micro-expressions and body movements. In all-reference tasks, characters, props, sounds, spatial relationships, and styles remain stable. The aesthetics of digital information displays, such as charts and UI content, have also improved.

Wan 3.0's video editing capabilities have been significantly enhanced, supporting modification of visuals, plots, and dialogue. However, there is still room for improvement in sound quality and text accuracy.

Starting immediately, Wan 3.0 is available for public beta on Alibaba Cloud Bailian, Wanjing Yike, Wanxiang official website, Qianwen Creation PC, IF STUDIO, and Duiyou, with grayscale access on the Qianwen app. API prices are 0.3/0.6/1.2 yuan per second for 480P/720P/1080P, and the API interface will be fully released soon.