← Back to community benchmarks

Qwen3.5-9B

M3 Max (40c) · 128 GB · 4bit · 2026-03-14
Performance
4k
tokens
738.5
PP tok/s
55.2
TG tok/s
5546
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.2.11
macOS macOS 15.7.5
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 748.2 67.4 6.7 GB view
4k 738.5 55.2 7.0 GB current
8k 723.7 53.7 7.2 GB view
16k 650.0 42.0 7.6 GB view
Batching Results
Batch Size TG tok/s Speedup
67.4 1.00×
60.6 0.90×
119.5 1.77×