← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M5 Pro (20c) · 64 GB · 4bit · 2026-05-14

Performance

8k

tokens

355.6

PP tok/s

19.7

TG tok/s

23040

TTFT (ms)

19.2

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

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	356.9	20.9	16.7 GB	view
4k	377.0	19.9	18.2 GB	view
8k	355.6	19.7	19.2 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	20.9	1.00×
2×	24.7	1.18×
4×	33.4	1.60×