← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M3 Max (40c) · 128 GB · 4bit · 2026-05-10
User commentNATIVE MTP ON
Performance
16k
tokens
133.8
PP tok/s
8.3
TG tok/s
122430
TTFT (ms)
20.7
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 15.7.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 189.2 21.2 16.8 GB view
4k 197.8 26.9 18.2 GB view
8k 194.2 23.2 19.2 GB view
16k 133.8 8.3 20.7 GB current
Batching Results
Batch Size TG tok/s Speedup
21.2 1.00×
12.2 0.58×
54.3 2.56×
70.5 3.33×