Alibaba’s Qwen3.7-Max Becomes World’s Second-Best AI Coding Model
Alibaba’s Qwen3.7-Max has secured the second position on Code Arena’s global AI coding leaderboard. As a result, the model now ranks above several systems developed by OpenAI, Google, and other major AI labs. The achievement also makes it the highest-ranked non-US model on the platform, highlighting the growing competition between Chinese and American AI developers.
According to rankings updated on May 26, the model scored 1541 in blind evaluations and trailed only Anthropic’s Claude models. Meanwhile, it outperformed OpenAI’s GPT-5.5, Google’s Gemini-3.5-Flash, Zhipu’s GLM-5.1, and Moonshot’s Kimi-K2.6.
Alibaba Cloud described the system as “officially the #2 AI coding model globally” based on Code Arena’s blind testing process, where judges review outputs without knowing which AI generated the code.
Blind Testing Measures Real-World Coding Skills
Code Arena uses randomized and anonymous comparisons to reduce brand bias during evaluations. In addition, the platform tests models across tasks such as web development, animation, game creation, and data visualization. Therefore, the rankings aim to reflect practical coding performance rather than isolated benchmark scores.
Qwen3.7-Max builds on Alibaba’s rapid progress in AI development over recent months. The company introduced the model during the Alibaba Cloud Summit in mid-May. Moreover, the reasoning-focused system includes a one-million-token context window designed for long-horizon coding, debugging, and agent-based workloads.
The model also scored 56.6 on the Artificial Analysis Intelligence Index, where it currently ranks fifth overall.
Chinese AI Firms Continue Challenging US Leaders
Alibaba had already gained momentum earlier this year. In March, the company’s Qwen3.5 medium models entered the top ten among open models in Code Arena rankings. Later, Qwen3.6-Plus attracted attention after outperforming Claude 4.5 Opus on several agentic coding benchmarks.
Although US firms still dominate much of the AI coding market, Chinese companies continue narrowing the gap. Alongside Alibaba, developers such as DeepSeek, Zhipu, and Moonshot are also delivering stronger performances across advanced coding evaluations.
With Qwen3.7-Max now available through Alibaba Cloud’s Model Studio API, the company is positioning itself as a stronger alternative for developers seeking advanced coding assistance.
