← Back to community benchmarks

Qwen3.5-9B

M5 Max (40c) · 128 GB · 4bit · 2026-03-13
Performance
4k
tokens
3,063
PP tok/s
97.2
TG tok/s
1337
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.2.10
macOS macOS 26.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,927 99.1 6.7 GB view
4k 3,063 97.2 7.0 GB current
8k 2,805 92.0 7.1 GB view
16k 2,524 81.2 7.6 GB view
32k 2,041 68.1 8.7 GB view
64k 1,626 38.8 10.8 GB view
Batching Results
Batch Size TG tok/s Speedup
99.1 1.00×
14.0 0.14×
61.1 0.62×
193.4 1.95×