Comment

Comment by Casteil

I don't know what black magic you're up to but I see more like 30-35t/s on a 16" M5 Max using 3.8:27b Q4, regardless of whether it's mlx or gguf.

qwen3.5:122b-a10b is significantly faster at around 60-65.

parent Casteil 2 months ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton