← Back to community benchmarks

Qwen3.6-27B-MTPLX-Optimized

M5 Max (40c) · 128 GB · 4bit · 2026-05-17

Performance

4k

tokens

914.5

PP tok/s

27.0

TG tok/s

4479

TTFT (ms)

21.1

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.9.dev2

macOS macOS 26.4.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	790.3	27.4	19.6 GB	view
4k	914.5	27.0	21.1 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	27.4	1.00×
2×	49.5	1.81×
4×	68.4	2.50×