Chinese e-commerce giant Alibaba has made its advanced artificial intelligence (AI) model, Wan 2.1, freely available to the public, marking a significant step in the global AI competition.

The open-source release of Wan 2.1—a tool capable of generating videos and images from text and image inputs—is poised to accelerate adoption and challenge rivals in the rapidly evolving AI landscape.

READ MORE: Alibaba commits over $52B to AI, cloud computing in next 3 years

The move mirrors a trend set earlier this year by startup DeepSeek, whose cost-effective open-source models stunned the industry by rivaling the performance of heavyweights like OpenAI, sparking investor enthusiasm.

Alibaba’s Wan 2.1 comes in four variants: T2V-1.3B, T2V-14B, I2V-14B-720P, and I2V-14B-480P. The "14B" designation highlights models with 14 billion parameters, enabling them to process vast amounts of data for more precise and realistic outputs.

These tools are now accessible worldwide via Alibaba Cloud’s ModelScope and HuggingFace platforms, catering to academic, research, and commercial users.

First unveiled in January as an upgrade to its Wanx model (later rebranded to Wan), Wan 2.1 has been celebrated for its ability to produce lifelike visuals. Alibaba has since touted its dominance on VBench, a leading benchmark for video generation, where it excels in capabilities like multi-object interaction.

This open-source release is expected to amplify its reach and influence.

In addition to Wan 2.1, Alibaba previewed its reasoning model, QwQ-Max, on Tuesday, with plans to make it open source upon its official launch. The company also revealed a massive investment of at least 380 billion yuan ($52 billion) over the next three years to enhance its cloud computing and AI infrastructure, underscoring its ambition to lead the sector.

‍