← Back to community benchmarks

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

M4 Max (40c) · 128 GB · 4bit · 2026-05-12
Performance
4k
tokens
251.5
PP tok/s
28.3
TG tok/s
16285
TTFT (ms)
17.3
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 243.7 29.3 15.8 GB view
4k 251.5 28.3 17.3 GB current
32k 229.1 24.4 21.6 GB view
Batching Results
Batch Size TG tok/s Speedup
29.3 1.00×
54.6 1.86×