← Back to community benchmarks

Qwen3.6-35B-A3B-OptiQ

M5 Pro (16c) · 48 GB · 4bit · 2026-07-23

Performance

16k

tokens

1,834

PP tok/s

63.2

TG tok/s

8933

TTFT (ms)

23.8

Peak mem (GB)

Hardware

Chip M5 Pro (16c)

Memory 48 GB

GPU Cores 16

Software

oMLX v0.5.2

macOS macOS 26.3.2

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	1,351	67.9	22.6 GB	view
8k	1,985	67.4	23.1 GB	view
16k	1,834	63.2	23.8 GB	current
32k	1,547	55.5	25.3 GB	view
64k	669.3	43.4	27.1 GB	view