← Back to community benchmarks

Qwen3.5-397B-A17Bvlm

M3 Ultra (80c) · 512 GB · 4bit · 2026-04-30

Performance

32k

tokens

473.2

PP tok/s

31.6

TG tok/s

69251

TTFT (ms)

216.0

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.3.8

macOS macOS 26.4.1

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	457.3	38.4	209.6 GB	view
8k	545.8	36.6	211.9 GB	view
32k	473.2	31.6	216.0 GB	current