← Back to community benchmarks

Qwen3.6-27B-MTPLX-Optimized-Speed

M3 Pro (18c) · 36 GB · 4bit · 2026-05-05

Performance

4k

tokens

94.9

PP tok/s

8.3

TG tok/s

43166

TTFT (ms)

17.3

Peak mem (GB)

Hardware

Chip M3 Pro (18c)

Memory 36 GB

GPU Cores 18

Software

oMLX v0.3.8

macOS macOS 15.7.5

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	89.7	8.3	15.8 GB	view
4k	94.9	8.3	17.3 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	8.3	1.00×
2×	16.4	1.98×
4×	21.7	2.61×