[AINews] Qwen 3.8 Max(2.4T) and 27B, new open weights models for Coding and Cowork
Latent.Space
Read full postAlibaba's Qwen lab is releasing open weights for Qwen 3.8 Max, a 2.4 trillion parameter model excelling in autonomous coding, AI research, hardware design, and multimodal tasks. The model demonstrates advanced capabilities including long-horizon autonomous coding, outperforming humans in data science challenges, and integrating visual feedback for multimodal agents.

- Qwen 3.8 Max· 5 sources
- Qwen3.8-Max: A New Bar for Coding and Cowork· Hacker News
- Qwen 3.8 Max now available on Vercel AI Gateway· Vercel


