← Back to community benchmarks

Qwen3.6-35B-A3B

M5 (10c) · 32 GB · 4bit · 2026-07-20

Performance

64k

tokens

489.4

PP tok/s

27.0

TG tok/s

133900

TTFT (ms)

22.8

Peak mem (GB)

Hardware

Chip M5 (10c)

Memory 32 GB

GPU Cores 10

Software

oMLX v0.5.1

macOS macOS 26.5.2

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,076	57.6	19.2 GB	view
4k	1,269	55.2	19.9 GB	view
8k	1,208	51.6	20.4 GB	view
16k	1,047	48.0	21.0 GB	view
32k	780.6	37.9	22.1 GB	view
64k	489.4	27.0	22.8 GB	current