← Back to community benchmarks

Qwen3.5-122B-A10B-VisionMixed

M5 Max (40c) · 128 GB · 4bit · 2026-04-01

Performance

8k

tokens

557.3

PP tok/s

16.0

TG tok/s

14699

TTFT (ms)

75.3

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.0

macOS macOS 26.4

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	393.3	17.1	73.3 GB	view
4k	549.2	16.8	74.7 GB	view
8k	557.3	16.0	75.3 GB	current
16k	522.5	14.9	76.5 GB	view
32k	454.6	13.5	78.9 GB	view
64k	335.5	11.1	83.7 GB	view