← Back to community benchmarks

Qwen3.6-35B-A3B-oQ4

M4 Max (40c) · 128 GB · 4bit · 2026-04-20

Performance

8k

tokens

1,272

PP tok/s

290.0

TG tok/s

6441

TTFT (ms)

28.8

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.6

macOS macOS 26.4.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,443	377.6	21.3 GB	view
4k	1,487	318.6	24.9 GB	view
8k	1,272	290.0	28.8 GB	current
16k	992.5	222.2	37.5 GB	view
32k	1,091	76.1	23.0 GB	view
64k	872.7	57.5	25.7 GB	view
128k	517.4	35.7	31.1 GB	view
195k	357.3	25.6	36.7 GB	view