← Back to community benchmarks

Qwen3.5-122B-A10B

M3 Ultra (80c) · 512 GB · 4bit · 2026-03-15

Performance

8k

tokens

908.7

PP tok/s

47.5

TG tok/s

9015

TTFT (ms)

67.0

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.2.12

macOS macOS 15.6

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
8k	908.7	47.5	67.0 GB	current
16k	851.5	43.7	67.7 GB	view