Hacker News new | ask | show | jobs
by scotty79 39 days ago
Qwen3.6-35B-A3B is pretty amazing. I'm using it with 96k context on 24GB VRAM through ollama.