← Back to community benchmarks

Qwen3.5-9B

M2 (8c) · 24 GB · 4bit · 2026-03-12
Performance
4k
tokens
84.6
PP tok/s
12.7
TG tok/s
48394
TTFT (ms)
6.8
Peak mem (GB)
Hardware
Chip M2 (8c)
Memory 24 GB
GPU Cores 8
Software
oMLX v0.2.8
macOS macOS 26.2
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 86.4 17.5 6.5 GB view
4k 84.6 12.7 6.8 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 17.5 1.00×
2× 12.2 0.70×
4× 13.7 0.78×