← Back to community benchmarks

Qwen3.5-27BMTP

M5 Max (40c) · 128 GB · 4bit · 2026-05-17
Performance
16k
tokens
695.5
PP tok/s
28.8
TG tok/s
23559
TTFT (ms)
19.8
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 786.8 32.9 15.8 GB view
4k 829.8 32.1 17.3 GB view
8k 771.8 29.6 18.3 GB view
16k 695.5 28.8 19.8 GB current
32k 659.3 28.3 22.8 GB view
Batching Results
Batch Size TG tok/s Speedup
32.9 1.00×
55.9 1.70×
70.5 2.14×
77.8 2.36×