← Back to community benchmarks

Qwen3.5-9B

M4 (10c) · 16 GB · 4bit · 2026-03-10
Performance
16k
tokens
142.1
PP tok/s
15.0
TG tok/s
115298
TTFT (ms)
7.5
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 152.9 18.6 6.5 GB view
4k 150.6 17.6 6.8 GB view
8k 149.7 15.8 7.0 GB view
16k 142.1 15.0 7.5 GB current
Batching Results
Batch Size TG tok/s Speedup
18.6 1.00×
28.4 1.53×
33.0 1.77×