Chinese e-commerce giant Alibaba has made its video and image-generating AI model, Wan 2.1, publicly available. This move paves the way for mass deployment and intensifies competition in the AI industry.
AI Model Variants and Availability
Open-sourcing AI models is a common trend, with DeepSeek being a notable player. Alibaba has introduced four variants of Wan 2.1: T2V-1.3B, T2V-14B, I2V-14B-720P, and I2V-14B-480P. These models generate videos and images from text queries or templates, including images. The numbers in their names indicate their parameter sizes—1.3 billion and 14 billion, respectively.
Users worldwide can access these models through Hugging Face and ModelScope (a part of Alibaba Cloud) for academic, research, and commercial purposes. Alibaba initially launched its AI-powered video generator in January under the name Wanx before renaming it Wan. The model performed well in Vbench tests for video generators, particularly excelling in object interaction criteria.
Future Investments in AI and Cloud Computing
Alongside Wan 2.1, Alibaba also released a preliminary version of the QwQ-Max reasoning model, which will be available as an open-source project, adds NIX Solutions. The company plans to invest at least 380 billion yuan ($52 billion) over the next three years to strengthen its cloud computing and AI infrastructure.
As AI development accelerates, Alibaba’s latest move underscores its commitment to advancing the field. We’ll keep you updated on further developments in AI and open-source releases.