← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M5 Pro (20c) · 48 GB · 4bit · 2026-05-17
Performance
64k
tokens
338.8
PP tok/s
18.0
TG tok/s
193418
TTFT (ms)
29.8
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 48 GB
GPU Cores 20
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 65,536
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 430.0 26.2 16.8 GB view
4k 466.0 24.5 18.2 GB view
8k 453.5 24.8 19.2 GB view
16k 437.7 23.8 20.7 GB view
32k 410.1 21.7 23.7 GB view
64k 338.8 18.0 29.8 GB current
Batching Results
Batch Size TG tok/s Speedup
26.2 1.00×
31.2 1.19×
41.5 1.58×