Why are the models being used so old (in llm years anyway). Why no Qwen 3.8 ?
Based on how vibe-coded this whole thing is, I’m guessing the newer models weren’t in the training data and the agent just picked what it knew existed.
Qwen 3.6 & 3.8 have the exact same model architecture. In theory, it should just be a drop-in replacement.
https://hfviewer.com/compare/qwen3.6-27b-vs-qwen3.8-27b
Based on how vibe-coded this whole thing is, I’m guessing the newer models weren’t in the training data and the agent just picked what it knew existed.
Qwen 3.6 & 3.8 have the exact same model architecture. In theory, it should just be a drop-in replacement.
https://hfviewer.com/compare/qwen3.6-27b-vs-qwen3.8-27b