← Back to community benchmarks

Nemotron-3-Nano-30B-A3B-TurboQuant

M5 Pro (20c) · 64 GB · 8bit · 2026-05-04
Performance
64k
tokens
4,057
PP tok/s
63.8
TG tok/s
16155
TTFT (ms)
34.9
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 65,536
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,372 67.4 33.7 GB view
4k 1,812 63.5 34.8 GB view
8k 1,894 64.6 34.9 GB view
32k 4,991 65.4 34.8 GB view
64k 4,057 63.8 34.9 GB current
128k 2,728 60.2 37.8 GB view
195k 2,061 56.9 41.0 GB view
Batching Results
Batch Size TG tok/s Speedup
67.4 1.00×
89.2 1.32×
109.7 1.63×
218.7 3.24×