Qwen 3.8 27B open model rivals frontier, tops Hugging Face

Alibaba's Qwen 3.8 27B, an Apache 2 licensed 27B vision-capable LLM, scores 52 on the Artificial Analysis Intelligence Index, matching GPT-5.6 Luna (max) and one point behind GLM-5.2 (max) and DeepSeek V4 Pro 0813 (max). It became the #1 trending model on Hugging Face and surpassed 1M downloads.
How this story unfolded
3 weeks · 5 reports · 52 community posts · from Aug 3
- Aug 3
- Aug 14
- Aug 15
- Aug 16
- Aug 17
- Aug 18
- Aug 19
- Aug 21
- Aug 22
- Aug 23
- Aug 24
- Aug 25
- Aug 26
Qwen by email
Get an email when Qwen has news
No news that day, no email.
More stories today
- Salesforce stock jumps 14% on AI growth and Anthropic investment gain
- OpenAI highlights Codex-built npm library
- Theo praises new model in video
- GitHub Copilot app automates Dependabot PR triage
- Observability faces data storage crisis as AI grows