On September 2, Alibaba updated its flagship model Qwen3.8-Max, which has seen a significant performance boost compared to the old version. After specialized post-training for programming and professional office work, the new version of Qwen3.8-Max has enhanced overall performance. On CodeArena, a global authoritative third-party ranking for front-end programming capabilities, its score increased by 22 points to 1691 points, surpassing models such as Claude Opus5 and Kimi K3 to rank first overall. Meanwhile, CodeArena's updated model price-performance ratio ranking (Pareto frontier) shows that the new version of Qwen3.8-Max averages only $5 per million tokens, directly surpassing all other models with prices over $5. It is understood that Qwen3.8-Max is Alibaba's most powerful large language model to date, with a total of 240 billion parameters and support for 1 million context tokens. The updated Qwen3.8-Max has emerged with even stronger intelligent programming capabilities, making it more suitable for complex tasks, scientific research, and long-cycle tasks. Currently, the new Qwen3.8-Max model has been launched on the Qianwen AI platform to provide external API services
This article is provided by Alibaba and reproduced by QbitAI with permission, with opinions belonging to the original author.