Qwen3.8-27B open-weights model launches, tops Hugging Face

How this story unfolded

2 weeks · 5 reports · 53 community posts · 58 of 59 shown

1. Aug 14
   • 6 community posts
2. Aug 15
   • orcarouter/Qwen3.8-27B-Uncensored-FP8huggingface.co
   • 3 community posts
3. Aug 16
   • Qwen 3.8 27B is excellent, but it defaults to wildly overthinking thingssimonwillison.net
   • 2 community posts
4. Aug 17
   • 0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored-GGUFhuggingface.co
   • Qwen3.8-27B lands next to DeepSeek V4 and GPT-5.6 Luna Max on the Artificial Analysis Benchmark. You can now run a near frontier model with just a RTX 3090.
5. Aug 18
   • Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Indexsimonwillison.net
   • Qwen 27B is the "DeepSeek moment" for open source. It matches the closed-source state-of-the-art...
6. Aug 19
   • Qwen3.8-27B is live on Tinker today. It’s natively multimodal with images + video, has flexible...
7. Aug 20
   • 7 community posts
8. Aug 21
   • 6 community posts
9. Aug 22
   • 2 community posts
10. Aug 23
    • orcarouter/Qwen3.8-27B-Uncensoredhuggingface.co
    • 7 community posts
11. Aug 24
    • 3 community posts
12. Aug 25
    • Qwen3.8 27B is now available for both fine-tuning and Dedicated Model Inference Fine-tune it on...
13. Aug 26
    • 8 community posts
14. Aug 27
    • 2 community posts
15. Aug 28
    • 3 community posts