← Back to community benchmarks

Qwen3.5-122B-A10B-5bit

M4 Max (40c) · 128 GB · 5bit · 2026-04-04

Performance

8k

tokens

502.0

PP tok/s

40.1

TG tok/s

16320

TTFT (ms)

82.6

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.2

macOS macOS 26.4

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	453.7	45.7	80.2 GB	view
4k	508.2	44.0	82.1 GB	view
8k	502.0	40.1	82.6 GB	current
16k	471.9	34.4	83.7 GB	view
32k	415.8	26.1	85.8 GB	view