← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M3 Pro (18c) · 36 GB · 4bit · 2026-05-15
Performance
1k
tokens
95.9
PP tok/s
13.6
TG tok/s
10674
TTFT (ms)
16.7
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 36 GB
GPU Cores 18
Software
oMLX v0.3.9.dev2
macOS macOS 26.4.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 95.9 13.6 16.7 GB current
4k 96.4 13.7 18.2 GB view
8k 95.4 12.4 19.2 GB view
16k 93.6 11.1 20.7 GB view
Batching Results
Batch Size TG tok/s Speedup
13.6 1.00×
16.5 1.21×
22.3 1.64×
23.2 1.71×