China's open source AI has gained traction with global developers
Alibaba and Zhipu AI have announced new open source models to strengthen China's efforts to compete with the US in AI development. Alibaba's WAN2.2 is being promoted as the first large-scale video generation model using a Mixture (MOE) architecture in the open source space.
The WAN2.2 series includes models for generating video from text and images, supporting hybrid capabilities in advanced multimedia applications. The MOE architecture allows these models to use less computing power by splitting tasks across specialized subnetworks.
Zhipu, one of China's leading AI companies, has launched the GLM-4.5 and GLM-4.5-Air models with up to 355 billion parameters, built on a self-developed architecture. The GLM-4.5 model ranked third and first among open source models globally across 12 performance benchmarks.
China's open source ecosystem is expanding rapidly, with Zhipu's model accumulating over 40 million downloads, and Alibaba's Qwen series producing hundreds of derivatives. Industry momentum reflects strategic shifts towards wider adoption, increased efficiency and increased international reach.
Want to learn more about AI, technology, and digital diplomacy? If so, ask our Diplo chatbot!
