← Back to community benchmarks

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

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

Performance

1k

tokens

307.6

PP tok/s

17.0

TG tok/s

3329

TTFT (ms)

15.8

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.3.9rc1

macOS macOS 26.4.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	307.6	17.0	15.8 GB	current
4k	341.8	16.7	17.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	17.0	1.00×
2×	28.0	1.65×
4×	33.6	1.98×