← Back to community benchmarks

Qwen3.5-122B-A10B

M5 Max (40c) · 128 GB · 4bit · 2026-04-04

Performance

32k

tokens

1,120

PP tok/s

37.4

TG tok/s

29250

TTFT (ms)

71.2

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.2

macOS macOS 26.3.2

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	1,434	62.1	67.0 GB	view
8k	1,523	56.9	67.7 GB	view
16k	1,443	52.7	68.9 GB	view
32k	1,120	37.4	71.2 GB	current
64k	850.9	33.6	76.1 GB	view
128k	546.3	21.4	85.7 GB	view