← Back to community benchmarks

Qwen3.6-35B-A3B-oQ4

M1 Ultra (48c) · 128 GB · 4bit · 2026-04-17

Performance

16k

tokens

2,292

PP tok/s

37.7

TG tok/s

7150

TTFT (ms)

22.7

Peak mem (GB)

Hardware

Chip M1 Ultra (48c)

Memory 128 GB

GPU Cores 48

Software

oMLX v0.3.6

macOS macOS 26.4.1

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,001	67.5	20.4 GB	view
4k	1,383	63.2	21.2 GB	view
8k	1,387	63.0	21.6 GB	view
16k	2,292	37.7	22.7 GB	current
32k	2,220	37.1	23.4 GB	view
64k	1,832	35.7	25.2 GB	view