← Back to community benchmarks

Qwen3.6-27B-oQ8-mtp

M5 Max (40c) · 128 GB · 8bit · 2026-05-21
Performance
32k
tokens
628.3
PP tok/s
25.3
TG tok/s
52154
TTFT (ms)
35.8
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9
macOS macOS 26.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 702.1 29.2 28.8 GB view
4k 760.7 28.5 30.3 GB view
8k 740.3 28.6 31.3 GB view
16k 682.6 26.3 32.8 GB view
32k 628.3 25.3 35.8 GB current
64k 549.8 22.0 41.8 GB view
128k 410.9 18.3 54.2 GB view