← Back to community benchmarks

Qwen3.6-27B-oQ8-mtp

M4 Max (40c) · 48 GB · 8bit · 2026-05-17

Performance

8k

tokens

210.6

PP tok/s

25.5

TG tok/s

38906

TTFT (ms)

31.3

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 48 GB

GPU Cores 40

Software

oMLX v0.3.9.dev2

macOS macOS 26.4.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	237.7	26.9	28.8 GB	view
4k	238.9	26.3	30.3 GB	view
8k	210.6	25.5	31.3 GB	current
16k	89.2	8.2	32.8 GB	view