← Back to community benchmarks

Qwen3.5-9B

M4 (10c) · 16 GB · 4bit · 2026-04-17
Performance
32k
tokens
186.6
PP tok/s
15.5
TG tok/s
175567
TTFT (ms)
8.9
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.3.5
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 217.0 20.5 5.7 GB view
8k 221.1 19.2 6.7 GB view
16k 202.8 17.8 7.4 GB view
32k 186.6 15.5 8.9 GB current
Batching Results
Batch Size TG tok/s Speedup
20.5 1.00×
41.4 2.02×
50.3 2.45×
51.9 2.53×