← Back to community benchmarks

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

M4 Max (32c) · 36 GB · 4bit · 2026-06-04
Performance
8k
tokens
120.1
PP tok/s
13.8
TG tok/s
68207
TTFT (ms)
18.0
Peak mem (GB)
Hardware
Chip M4 Max (32c)
Memory 36 GB
GPU Cores 32
Software
oMLX v0.4.1
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 99.9 8.7 15.3 GB view
4k 83.5 9.2 16.3 GB view
8k 120.1 13.8 18.0 GB current
Batching Results
Batch Size TG tok/s Speedup
8.7 1.00×
22.9 2.63×
31.2 3.59×
35.0 4.02×