On August 3, the QianWen team of Alibaba Cloud unveiled the Qwen 3.8-Max large-scale model, which is equipped with a staggering 2.4 trillion parameters (95 billion active parameters). This model is primarily designed to enhance end-to-end processing capabilities across various domains, including programming, office tasks, scientific research, and long-term projects. Furthermore, Alibaba Cloud has announced plans to open-source the model's weights next week. As the most formidable model in the Qwen series, Qwen 3.8-Max is constructed on the foundation of the Qwen 3.5 architecture and is now accessible on the QianWen AI platform, providing API services. This move signifies the first instance where a Qwen-Max level model has been made open-source.