← Back to community benchmarks

Qwen3.5-9B

M4 Max (40c) · 64 GB · 4bit · 2026-03-13
Performance
8k
tokens
826.4
PP tok/s
80.0
TG tok/s
9912
TTFT (ms)
7.1
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 64 GB
GPU Cores 40
Software
oMLX v0.2.10
macOS macOS 26.3.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 885.0 88.7 6.7 GB view
4k 889.2 85.3 7.0 GB view
8k 826.4 80.0 7.1 GB current
Batching Results
Batch Size TG tok/s Speedup
88.7 1.00×
154.7 1.74×
173.9 1.96×
191.8 2.16×