On August 3, 2026, Alibaba proudly introduced its latest-generation foundational large model, Qwen3.8. This cutting-edge model boasts an impressive 2.4 trillion parameters and showcases significantly enhanced capabilities, particularly in programming and professional office applications. On the prestigious Arena leaderboard, the Qwen model stands tall among the world's top-performing large models. Currently, the API for Qwen3.8 is accessible on the QianWen AI platform and has been seamlessly integrated into the newly released 'QianWen Office' Agent product. Excitingly, Qwen3.8-Max and Qwen3.8-27B are slated for open-source release next week. Developers can leverage API services via the QianWen AI platform, with domestic pricing set at a competitive 12 yuan per million input tokens and 36 yuan per million output tokens. Moreover, implicit cache hits are priced at a mere 1.5 yuan, offering developers an economical and efficient solution.