← Back to community benchmarks

Qwen36-35B-A3B-oQ8

M4 Max (40c) · 128 GB · 8bit · 2026-04-30

Performance

16k

tokens

1,583

PP tok/s

76.5

TG tok/s

10349

TTFT (ms)

37.1

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.8

macOS macOS 15.7.4

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,209	244.9	36.1 GB	view
4k	873.0	84.4	36.2 GB	view
8k	1,684	81.7	36.5 GB	view
16k	1,583	76.5	37.1 GB	current
32k	1,375	70.5	38.5 GB	view