if on MacOS I recommend llm-mlx which currently renders tokens 10%-15% faster th... | Hacker News

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

shironnnn_ 59 days ago | parent | context | favorite | on: Ask HN: Has anyone replaced Claude/GPT with a loca...

if on MacOS I recommend llm-mlx which currently renders tokens 10%-15% faster than llama.cpp.

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