← Back to community benchmarks

Qwen3.5-9B-OptiQ

M3 Pro (18c) · 36 GB · 4bit · 2026-03-12
Performance
4k
tokens
344.2
PP tok/s
25.2
TG tok/s
11901
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 36 GB
GPU Cores 18
Software
oMLX v0.2.10
macOS macOS 26.0.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 324.2 25.9 6.8 GB view
4k 344.2 25.2 7.0 GB current
8k 341.9 24.2 7.2 GB view
16k 331.5 22.5 7.7 GB view
32k 309.9 20.0 8.8 GB view
64k 269.3 15.4 10.9 GB view
Batching Results
Batch Size TG tok/s Speedup
25.9 1.00×
51.2 1.98×
102.5 3.96×
204.8 7.91×