← Back to community benchmarks

Qwen3.6-27B-MTPLX-Optimized-Speed

M5 Pro (20c) · 64 GB · 4bit · 2026-06-17
Performance
1k
tokens
351.1
PP tok/s
18.3
TG tok/s
2917
TTFT (ms)
15.8
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.4.4
macOS macOS 27.0
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 351.1 18.3 15.8 GB current
4k 396.8 17.5 17.3 GB view
Batching Results
Batch Size TG tok/s Speedup
18.3 1.00×
31.0 1.69×
39.1 2.14×
34.1 1.86×