Alibaba just dropped the biggest AI model to come out of China, and the market noticed immediately. The company’s new Qwen3.8-Max-Preview, packing 2.4 trillion parameters, launched as a live preview over the weekend, with Alibaba claiming it trails only Anthropic’s Claude Fable 5 in overall performance.
Alibaba shares climbed as much as 5.4% in Hong Kong following the announcement.
What Alibaba actually built
The Qwen3.8-Max-Preview represents the latest evolution of Alibaba’s Qwen series, which the company first launched back in 2023. At 2.4 trillion parameters, this model sits comfortably among the largest language models ever released.
Alibaba has made the model available on its coding platforms, including Token Plan, and committed to releasing the open weights soon. Open weights mean other developers and researchers can build on top of the model, which is how you create an ecosystem rather than just a product.
