You can expect around 55-60t/s with Qwen3.5:35b-a3b or gemma4:26b-a4b Q4 | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

Casteil 78 days ago | parent | context | favorite | on: Running local models on an M4 with 24GB memory

You can expect around 55-60t/s with Qwen3.5:35b-a3b or gemma4:26b-a4b Q4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact