← Back to community benchmarks

Qwen3.5-VL-122B-A10B-4bitCRACK

M3 Ultra (80c) · 512 GB · 4bit · 2026-04-05

Performance

32k

tokens

783.4

PP tok/s

40.2

TG tok/s

41828

TTFT (ms)

71.1

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.3.4

macOS macOS 26.4

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
16k	897.7	47.0	68.7 GB	view
32k	783.4	40.2	71.1 GB	current