← Back to community benchmarks

Qwen3.6-27B-MXFP4-MTP

M4 Max (32c) · 36 GB · 4bit · 2026-05-15

Performance

16k

tokens

187.6

PP tok/s

29.3

TG tok/s

87333

TTFT (ms)

20.5

Peak mem (GB)

Hardware

Chip M4 Max (32c)

Memory 36 GB

GPU Cores 32

Software

oMLX v0.3.8

macOS macOS 15.3

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
8k	193.3	31.4	18.1 GB	view
16k	187.6	29.3	20.5 GB	current
32k	172.6	26.2	23.6 GB	view