Hacker News new | ask | show | jobs
by supjeff 41 days ago
I'm having some success running `qwen/qwen3.6-35b-a3b` in LMStudio with Opencode. I'm on an M4 MBP with 36gb. I'm getting 80 tok/sec with 260k context limit and temperature set to 0 (same prompt results in same output every time).