← Back to community benchmarks

Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated

M5 Max (40c) · 128 GB · 6bit · 2026-04-26

Performance

8k

tokens

3,707

PP tok/s

84.3

TG tok/s

2210

TTFT (ms)

28.4

Peak mem (GB)

Hardware

Chip M5 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	1,791	92.1	27.3 GB	view
4k	3,190	86.9	28.1 GB	view
8k	3,707	84.3	28.4 GB	current
16k	3,475	87.6	29.1 GB	view
32k	2,762	64.7	30.4 GB	view