← Back to community benchmarks

Qwen3.6-35B-A3B-UD

M4 Max (40c) · 128 GB · 4bit · 2026-07-23

Performance

16k

tokens

772.9

PP tok/s

67.1

TG tok/s

21197

TTFT (ms)

21.7

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.5.3

macOS macOS 26.5.1

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	730.7	68.4	21.2 GB	view
8k	778.1	67.3	21.4 GB	view
16k	772.9	67.1	21.7 GB	current
32k	592.7	38.1	22.2 GB	view
64k	288.6	17.1	23.3 GB	view
128k	230.5	14.7	25.6 GB	view