← Back to community benchmarks

Qwen3.6-27B

M5 Max (40c) · 128 GB · 8bit · 2026-07-23

Performance

64k

tokens

470.5

PP tok/s

14.5

TG tok/s

139283

TTFT (ms)

41.4

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.5.2.dev1

macOS macOS 26.5.2

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	786.6	16.7	28.4 GB	view
4k	813.7	16.7	29.8 GB	view
8k	690.0	15.7	30.9 GB	view
16k	606.1	15.6	32.4 GB	view
32k	567.6	15.4	35.4 GB	view
64k	470.5	14.5	41.4 GB	current
128k	367.4	12.8	53.8 GB	view