← Back to community benchmarks

Qwen3.6-35B-A3Bmxfp8

M4 Max (40c) · 128 GB · 8bit · 2026-04-16

Performance

32k

tokens

1,235

PP tok/s

59.5

TG tok/s

26524

TTFT (ms)

37.5

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.5

macOS macOS 26.2

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
16k	1,394	69.9	36.1 GB	view
32k	1,235	59.5	37.5 GB	current
64k	913.5	45.4	40.2 GB	view