← Back to community benchmarks

Qwen3.6-27B-oQ6-mtp

M5 Max (40c) · 128 GB · 6bit · 2026-05-22
Performance
32k
tokens
616.9
PP tok/s
26.1
TG tok/s
53122
TTFT (ms)
29.9
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 701.1 33.1 23.0 GB view
4k 816.6 32.8 24.4 GB view
8k 779.7 31.4 25.4 GB view
16k 688.3 28.9 26.9 GB view
32k 616.9 26.1 29.9 GB current
64k 530.7 22.2 36.0 GB view
128k 382.8 16.7 48.3 GB view
Batching Results
Batch Size TG tok/s Speedup
33.1 1.00×
34.6 1.05×
40.6 1.23×
83.5 2.52×