← Back to community benchmarks

Qwen3.5-9B4.5bit

M4 (10c) · 16 GB · 4bit · 2026-03-12
Performance
4k
tokens
222.2
PP tok/s
20.1
TG tok/s
18435
TTFT (ms)
5.9
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 4,096