Alibaba's Qwen team has announced Qwen 3.8 Max, a new 2.4T-parameter flagship model focused on coding, long-horizon agentic work, and multimodal reasoning. The company confirmed that open-weight versions of both Qwen 3.8 Max and the smaller Qwen 3.8-27B will be released next week.
- The model features a 1M-token context window, 95B active parameters per token (approx. 4% activation ratio), and native visual feedback in its execution loop.
- Alibaba claims capabilities including 10+ days of autonomous coding, 500+ turns of chip design optimization, and 365 days of e-commerce strategy execution.
- Independent evaluations place Qwen 3.8 Max at #4 overall in Frontend Code Arena with 1,668 Elo and #2 in Vision Arena with 1,305 points.
- On the Vals Index, it ranks #2 among open-weight models with a score of 66.1, matching Claude Opus 4.7 at a lower cost per test ($2.68 vs $6.17).
- API pricing is set at $2.00 per million input tokens and $6.00 per million output tokens.
The announcement signals that the Chinese open-weight frontier is now competing directly with top Western closed models, particularly in coding and agentic workflows.