Today we celebrate the launch and open sourcing of Qwen 3.8 27b which is a nice step forward for local AI.
For its size (~15GB on disk, 4bit quant) it is unmatched for agentic work on your personal computer. And it's ability to understand images and videos is refreshing. But...
Model: Nemotron-3.5-Lightning-30B-A3B-NVFP4
Harness: Hermes Agents Thoughts compared to Qwen 3.8 27B - NVFP4
This is Alibaba's new open source model, Qwen 3.8 (27B), which came into our lives yesterday.
I can clearly say that Alibaba has done a really good job. I was really impressed by the performance of the model and wanted to share my preliminary experiences with you.
Qwen 3.8 27b is my favorite release this year by far, way better release than any frontier model.
Releases could stop forever and I'd be happy! The 0 shots are so good lol.
You have reached the end of the archive
All of qwen38