← Back to community benchmarks

Qwen3.5-4B

M2 Pro (16c) · 16 GB · 4bit · 2026-03-11
Performance
16k
tokens
302.8
PP tok/s
40.7
TG tok/s
54110
TTFT (ms)
5.3
Peak mem (GB)
Hardware
Chip M2 Pro (16c)
Memory 16 GB
GPU Cores 16
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 323.2 59.8 4.0 GB view
4k 319.7 54.8 4.2 GB view
8k 312.2 49.6 4.6 GB view
16k 302.8 40.7 5.3 GB current
32k 281.4 31.0 6.9 GB view
Batching Results
Batch Size TG tok/s Speedup
59.8 1.00×
68.7 1.15×
73.1 1.22×
63.8 1.07×