Qwen 3.8 27B open-weights model released, scores 52 on AI Index

Alibaba's Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM, outperforms Qwen3.7-Plus and scores 52 on the Artificial Analysis Intelligence Index, matching GPT-5.6 Luna. It defaults to xhigh reasoning effort, causing overthinking, and runs locally on consumer hardware.
How this story unfolded
10 days · 9 reports · 37 community posts · from Aug 14
- Aug 14
- Aug 15
- Aug 16
- Aug 17
- Aug 18
- Aug 19
- Aug 20
- Aug 21
- Aug 22
- Aug 23
- Aug 24
Qwen by email
Get an email when Qwen has news
No news that day, no email.
More stories today
- Mystery model Ox Alpha identified as GLM-5.3 Flash
- Debate over AI competence and regulation timing
- Apodex 1.1 scores 56.1 on Humanity's Last Exam, beating Claude Opus 5
- Robotic systems struggle with physical navigation in recent demonstrations
- ChatGPT Visualize skill turns notes into interactive interfaces