← Back to community benchmarks

Qwen3.5-9B

M3 Ultra (60c) · 96 GB · 4bit · 2026-03-09
Performance
4k
tokens
1,135
PP tok/s
102.7
TG tok/s
3609
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 96 GB
GPU Cores 60
Software
oMLX v0.2.6
macOS macOS 26.2
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,122 107.1 6.7 GB view
4k 1,135 102.7 7.0 GB current
Batching Results
Batch Size TG tok/s Speedup
107.1 1.00×
108.0 1.01×
153.4 1.43×
194.8 1.82×