← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M5 Pro (20c) · 48 GB · 4bit · 2026-05-24

Performance

32k

tokens

406.0

PP tok/s

21.9

TG tok/s

80707

TTFT (ms)

20.0

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 48 GB

GPU Cores 20

Software

oMLX v0.3.9

macOS macOS 26.4.1

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	427.5	26.4	16.2 GB	view
4k	465.9	26.2	16.4 GB	view
8k	463.1	24.9	17.3 GB	view
16k	440.4	24.2	18.1 GB	view
32k	406.0	21.9	20.0 GB	current
64k	339.3	17.7	23.9 GB	view
128k	251.3	13.9	31.6 GB	view