Qwen 3.8 27B matches frontier models, runs locally

How this story unfolded

3 weeks · 5 reports · 118 community posts · 123 of 133 shown

1. Aug 3
   • 2 community posts
2. Aug 13
   • 4 community posts
3. Aug 14
   • 21 community posts
4. Aug 15
   • 7 community posts
5. Aug 16
   • Qwen 3.8 27B is excellent, but it defaults to wildly overthinking thingssimonwillison.net
   • 8 community posts
6. Aug 17
   • 16 community posts
7. Aug 18
   • Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Indexsimonwillison.net
   • Qwen3.8-27B runs frontier-class coding agents and reasoning locally, no cloud API requiredventurebeat.com
   • 10 community posts
8. Aug 19
   • bartowski/Qwen3.8-27B-GGUFhuggingface.co
   • 6 community posts
9. Aug 20
   • 11 community posts
10. Aug 21
    • AI News: The Best Open Model Runs on Your Computer!youtube.com
    • 8 community posts
11. Aug 22
    • 4 community posts
12. Aug 23
    • 13 community posts
13. Aug 24
    • 6 community posts
14. Aug 25
    • 2 community posts