← Back to community benchmarks

Qwen3.5-9B

M4 (10c) · 32 GB · 8bit · 2026-04-25
Performance
16k
tokens
483.6
PP tok/s
9.4
TG tok/s
33880
TTFT (ms)
12.5
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.6
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 179.0 77.5 12.2 GB view
4k 177.7 11.6 11.4 GB view
8k 220.2 11.3 11.8 GB view
16k 483.6 9.4 12.5 GB current
32k 457.9 9.3 13.2 GB view