← Back to community benchmarks

Qwen3.6-35B-A3B-oQ8-fp16-mtp

M5 Max (40c) · 128 GB · 8bit · 2026-06-07

Performance

128k

tokens

1,212

PP tok/s

50.4

TG tok/s

108118

TTFT (ms)

48.7

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.4.1

macOS macOS 26.5

Context 131,072

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	1,890	68.9	37.0 GB	view
8k	2,068	64.0	37.5 GB	view
16k	2,072	55.2	38.2 GB	view
32k	1,907	54.2	39.7 GB	view
64k	1,477	48.3	42.7 GB	view
128k	1,212	50.4	48.7 GB	current