← Back to community benchmarks

NVIDIA-Nemotron-3-Nano-30B-A3B

M5 Pro (20c) · 64 GB · 8bit · 2026-04-26
Performance
8k
tokens
1,900
PP tok/s
64.6
TG tok/s
4313
TTFT (ms)
34.6
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,362 66.9 33.5 GB view
4k 1,852 66.2 34.6 GB view
8k 1,900 64.6 34.6 GB current
16k 1,893 58.3 34.6 GB view
32k 1,801 46.3 34.7 GB view
64k 1,569 41.8 34.9 GB view
Batching Results
Batch Size TG tok/s Speedup
66.9 1.00×
91.2 1.36×
112.7 1.68×
127.9 1.91×