← Back to community benchmarks

Qwen3.5-9B-9bit

M3 Ultra (60c) · 96 GB · 8bit · 2026-03-09
Performance
32k
tokens
986.5
PP tok/s
50.3
TG tok/s
33218
TTFT (ms)
12.4
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 96 GB
GPU Cores 60
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 877.8 65.4 10.4 GB view
4k 1,052 64.0 10.6 GB view
8k 1,070 62.3 10.8 GB view
16k 1,048 58.0 11.3 GB view
32k 986.5 50.3 12.4 GB current
Batching Results
Batch Size TG tok/s Speedup
65.4 1.00×
110.3 1.69×
144.1 2.20×