← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M5 Max (40c) · 128 GB · 4bit · 2026-05-19

Performance

32k

tokens

724.6

PP tok/s

27.6

TG tok/s

45220

TTFT (ms)

23.4

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.9.dev2

macOS macOS 26.5

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	769.4	31.8	16.4 GB	view
4k	898.4	31.2	17.9 GB	view
8k	869.2	30.6	18.9 GB	view
16k	800.1	29.3	20.4 GB	view
32k	724.6	27.6	23.4 GB	current