← Back to community benchmarks

Qwen3.5-4B

M4 Pro (16c) · 24 GB · unknown · 2026-03-11
Performance
8k
tokens
558.8
PP tok/s
2.9
TG tok/s
14660
TTFT (ms)
12.4
Peak mem (GB)
Hardware
Chip M4 Pro (16c)
Memory 24 GB
GPU Cores 16
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 499.2 5.6 12.0 GB view
4k 559.5 5.6 12.1 GB view
8k 558.8 2.9 12.4 GB current
Batching Results
Batch Size TG tok/s Speedup
5.6 1.00×
35.5 6.34×