← Back to community benchmarks

Qwen3.6-27B-oQ8-mtp

M4 Max (40c) · 128 GB · 8bit · 2026-05-22
Performance
8k
tokens
246.8
PP tok/s
16.0
TG tok/s
33197
TTFT (ms)
30.8
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9
macOS macOS 26.5
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 235.3 16.5 28.3 GB view
4k 248.5 16.3 29.8 GB view
8k 246.8 16.0 30.8 GB current
16k 239.6 15.5 32.3 GB view
32k 226.4 14.8 35.3 GB view
Batching Results
Batch Size TG tok/s Speedup
16.5 1.00×
31.7 1.92×
45.5 2.76×