← Back to community benchmarks

Qwen3.6-27B-oQ8e-mtp

M3 Ultra (80c) · 512 GB · 8bit · 2026-07-17

Performance

64k

tokens

334.1

PP tok/s

18.0

TG tok/s

196157

TTFT (ms)

36.1

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.5.1

macOS macOS 26.5.2

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	418.3	21.6	28.3 GB	view
4k	427.5	21.4	29.5 GB	view
8k	422.2	18.1	30.2 GB	view
16k	407.3	20.6	31.4 GB	view
32k	379.1	19.6	32.7 GB	view
64k	334.1	18.0	36.1 GB	current
128k	269.4	15.6	42.1 GB	view