Alibaba Cloud Launches Wan3.0 Video Model Supporting 30-Second Generation and Document Inputs
Alibaba Cloud has officially launched its Wan3.0 video generation model, which can generate native 30-second videos in a single take. The model also introduces the ability to accept various document formats, including DOC, XLS, PPT, PDF, and MD, as inputs for video creation. This release advances the generative AI video space by enabling longer native video outputs and allowing direct integration of complex document data into visual content. It lowers the barrier for businesses and creators to turn structured documents into engaging video presentations. The model is available on platforms like Alibaba Cloud Bailian and Tongyi Wanxiang, with API pricing set at 0.3, 0.6, and 1.2 RMB per second for 480p, 720p, and 1080p resolutions, respectively. It also supports up to 20 reference assets, including webpage parsing, to maintain consistency.
## BACKGROUND
Alibaba Cloud's Bailian Platform, launched in October 2023, is a comprehensive large model development and service platform that manages APIs and AI services. Wan3.0 is part of Alibaba's broader AI ecosystem, which includes the Qwen large language models and Tongyi Wanxiang visual models, designed to compete with global AI video generators.