Alibaba's Qwen large model officially released Qwen4-Max on November 16, positioning it as an around-the-clock AI collaboration partner. The model is based on the new Qwen4 architecture, with 2.4T total parameters, 71B activated per token, and an additional 328B N-gram Embedding parameters. It has native vision capabilities and supports a million-token context.
It is now available on the Qwen AI platform for API services, with cache hits as low as RMB 0.5 per million tokens. It can also be accessed via Alibaba Token Plan or experienced in applications such as Qwen Office and Qoder; official API pricing has been comprehensively lowered, and small and medium-sized models in the Qwen4 series will be open-sourced progressively.




