← Back to community benchmarks

Qwen3.5-122B-A10B

M4 Max (40c) · 128 GB · 4bit · 2026-03-19

Performance

32k

tokens

485.3

PP tok/s

34.3

TG tok/s

67518

TTFT (ms)

72.1

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.2.18

macOS macOS 26.3

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
8k	587.4	50.2	68.5 GB	view
16k	551.3	43.0	69.7 GB	view
32k	485.3	34.3	72.1 GB	current
64k	380.9	23.8	77.0 GB	view
128k	265.8	14.6	86.7 GB	view
195k	200.5	8.6	96.8 GB	view