On November 11, the official Qwen large model team teased that the next-generation Qwen large model, Qwen4, would soon be released, and that a Qwen4-Max preview was being opened at the same time. According to the team, the new model continues to advance in knowledge understanding, world perception, coding, and agentic capabilities, with deeper knowledge understanding and more multidimensional world perception; from complex programming and professional office work to front-end design and professional software operation, Qwen4 has already delivered initial results in real-world task tests, but it did not disclose specific parameters, benchmark scores, or an official release date.
The Qwen4-Max preview made its debut on Alibaba's Token Plan, where users can try it for a limited time at as low as one-tenth of the original price. Access is through the Token Plan page on the Qwen AI platform. In-house products such as Qwen Office, Qoder, the Qwen App, and QwenStudio were among the first to integrate the new model, creating synchronized coverage from the model layer to the application layer.
This combined move—launching a preview first, immediately integrating it across the entire product line, and offering low-cost access—places the focus of validation on real-world task feedback rather than parameter disclosures. It continues the industry trend in which competition among domestic large models shifts from launch metrics to developer ecosystems and actual task performance.
