On August 3, 2026, Alibaba made an official announcement regarding the launch of its cutting-edge foundational large model, Qwen3.8-Max. This new model is a technological marvel, boasting an impressive 2.4 trillion parameters. It represents a substantial leap forward, particularly in the realms of programming and professional office functionalities.

When evaluated on the authoritative third-party benchmark Arena, the Qwen model has demonstrated remarkable performance. It stands as a strong contender, ranking second only to Anthropic's renowned Claude series. In the broader context of global large models, its overall performance places it firmly in the top echelon.

At present, developers can access the API for Qwen3.8 through the QianWen AI platform. Moreover, this API has been seamlessly integrated with the newly launched Agent product, 'QianWen Office,' offering a comprehensive and efficient solution for various office-related tasks.

Looking ahead, there is exciting news for the global developer community. The Qwen3.8-Max and Qwen3.8-27B models are slated to be open-sourced next week. This move is set to provide developers worldwide with a robust, high-performance, and cost-effective foundation for AI computing, fostering innovation and collaboration on a global scale.