← Back to community benchmarks

Qwen3.6-27B-mtp

M5 Max (40c) · 128 GB · 4bit · 2026-05-16
Performance
4k
tokens
898.4
PP tok/s
47.5
TG tok/s
4559
TTFT (ms)
17.6
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 756.3 45.4 16.1 GB view
4k 898.4 47.5 17.6 GB current
8k 832.3 46.8 18.6 GB view
16k 700.4 40.7 20.1 GB view
Batching Results
Batch Size TG tok/s Speedup
45.4 1.00×
55.6 1.22×
70.9 1.56×