On the morning of May 26, 2026, Code Arena—a globally recognized third-party programming leaderboard—published its latest rankings. Alibaba’s cutting-edge flagship model, Qwen3.7-Max, outperformed competitors like GPT-5.5 and Gemini-3.5-Flash, securing a remarkable score of 1541. This achievement placed it second among large-scale AI model providers worldwide, trailing only the Claude series, and marked it as the sole domestic model to surpass the 1540-point threshold. Built specifically for agent-based applications, Qwen3.7-Max demonstrates groundbreaking advancements in core capabilities, including programming, intelligent automation, and sustained task execution. The model is capable of independently managing end-to-end development of complex projects, operating continuously for 35 hours with over 1,000 cumulative tool invocations, and even self-optimizing chip kernels through autonomous programming.