← Back to community benchmarks

Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved-source-oQ4-mtp

M4 Max (40c) · 128 GB · 4bit · 2026-05-21

Performance

8k

tokens

183.3

PP tok/s

26.9

TG tok/s

44684

TTFT (ms)

18.8

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.9rc1

macOS macOS 26.4.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	229.4	37.8	16.8 GB	view
8k	183.3	26.9	18.8 GB	current
16k	175.6	24.2	20.7 GB	view
32k	122.7	6.1	23.7 GB	view