← Back to community benchmarks

Qwen3.6-27B-mtp

M3 Max (40c) · 128 GB · 8bit · 2026-05-08
Performance
4k
tokens
209.2
PP tok/s
19.7
TG tok/s
19583
TTFT (ms)
30.3
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 204.9 21.0 28.8 GB view
4k 209.2 19.7 30.3 GB current
8k 186.6 19.3 31.3 GB view
Batching Results
Batch Size TG tok/s Speedup
1× 21.0 1.00×
2× 24.2 1.15×
4× 33.2 1.58×