Qwen 3.8 Flash Next

Software / App

An LLM model mentioned as utilizing GQA with G=2 and H=24 for memory efficiency.

Mentioned in 1 video