← Back to community benchmarks

Qwen3.5-4B-OptiQ

M4 Pro (20c) · 48 GB · 4bit · 2026-04-25
Performance
32k
tokens
647.8
PP tok/s
41.8
TG tok/s
50586
TTFT (ms)
7.2
Peak mem (GB)
Hardware
Chip M4 Pro (20c)
Memory 48 GB
GPU Cores 20
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 711.4 69.0 4.0 GB view
4k 775.8 66.0 4.6 GB view
8k 767.9 62.1 5.0 GB view
16k 730.5 55.2 5.7 GB view
32k 647.8 41.8 7.2 GB current
64k 529.1 30.6 10.3 GB view
128k 370.5 16.5 16.6 GB view
195k 274.0 11.8 23.2 GB view
Batching Results
Batch Size TG tok/s Speedup
69.0 1.00×
119.3 1.73×
151.2 2.19×
160.5 2.33×