← Back to community benchmarks

Qwen3.5-4B-OptiQ

M4 (10c) · 16 GB · 4bit · 2026-03-10
Performance
1k
tokens
196.4
PP tok/s
19.2
TG tok/s
5214
TTFT (ms)
4.8
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 196.4 19.2 4.8 GB current
4k 209.0 19.4 5.4 GB view
8k 201.8 17.7 5.7 GB view
16k 191.7 13.5 6.7 GB view
Batching Results
Batch Size TG tok/s Speedup
19.2 1.00×
33.9 1.77×
38.7 2.02×