← Back to community benchmarks

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

M5 Pro (20c) · 64 GB · 8bit · 2026-04-27
Performance
1k
tokens
1,360
PP tok/s
67.4
TG tok/s
753
TTFT (ms)
49.1
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 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,360 67.4 49.1 GB current
4k 1,786 66.0 50.1 GB view
8k 1,834 64.1 50.1 GB view
16k 1,537 65.3 50.2 GB view
32k 1,468 64.4 50.3 GB view
64k 1,340 62.8 50.8 GB view
Batching Results
Batch Size TG tok/s Speedup
67.4 1.00×
90.5 1.34×
111.3 1.65×
125.4 1.86×