← Back to community benchmarks

Qwen3.5-9B

M3 Pro (18c) · 18 GB · 4bit · 2026-03-10
Performance
4k
tokens
347.4
PP tok/s
25.0
TG tok/s
11790
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 18 GB
GPU Cores 18
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 336.2 25.3 6.7 GB view
4k 347.4 25.0 7.0 GB current
8k 342.7 24.3 7.1 GB view
16k 330.3 22.6 7.6 GB view
32k 305.7 19.1 8.7 GB view
Batching Results
Batch Size TG tok/s Speedup
25.3 1.00×
51.6 2.04×
81.1 3.21×
83.2 3.29×