← Back to community benchmarks

Huihui-Qwen3.6-27B-abliterated

M5 Max (40c) · 128 GB · bf16 · 2026-04-26

Performance

16k

tokens

445.2

PP tok/s

65.3

TG tok/s

36798

TTFT (ms)

59.8

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 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	810.2	76.2	55.6 GB	view
4k	772.0	74.5	57.9 GB	view
8k	635.2	71.3	59.5 GB	view
16k	445.2	65.3	59.8 GB	current
32k	296.1	52.7	62.7 GB	view