← Back to community benchmarks

Qwen3.5-4B-OptiQ

M5 Pro (16c) · 48 GB · 4bit · 2026-04-18
Performance
1k
tokens
1,619
PP tok/s
80.3
TG tok/s
633
TTFT (ms)
4.0
Peak mem (GB)
Hardware
Chip M5 Pro (16c)
Memory 48 GB
GPU Cores 16
Software
oMLX v0.3.6
macOS macOS 26.4
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,619 80.3 4.0 GB current
64k 1,322 38.2 10.3 GB view