← Back to community benchmarks

Qwen3.6-27B-OptiQ

M5 Max (40c) · 128 GB · 4bit · 2026-06-19

Performance

32k

tokens

723.7

PP tok/s

25.0

TG tok/s

45276

TTFT (ms)

25.0

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.4.4rc1

macOS macOS

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
32k	723.7	25.0	25.0 GB	current
64k	538.0	21.2	32.2 GB	view
128k	339.5	15.5	44.6 GB	view