AI TRENDS | Alibaba Updates Qwen3.8-Max, Tops CodeArena With 1,691 Points

According to Wallstreetcn, Alibaba updated its flagship Qwen3.8-Max model on September 2, with performance significantly improved over the previous version after post-training focused on coding and professional office tasks. In the CodeArena benchmark for front-end coding, Qwen3.8-Max rose 22 points to 1,691, ranking first overall ahead of Claude Opus5 and Kimi K3, while the updated model's average cost was $5 per million tokens on CodeArena's value-for-money list.