Model benchmarks
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e local LLM performance
As of August 2026, NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e runs at up to 106.6 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
oMLX
Model size
30B
Peak speed
106.6 tok/s
Average speed
106.5 tok/s
Min memory
31.8 GB
Max context
27,108 tokens
Best quality
89.1
Benchmark runs
2
GPUs tested
1
Performance by hardware and tool
Every hardware/tool/quantization combination NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Apple M5 Max | oMLX | — | 106.6 tok/s | 106.5 tok/s | 31.8 GB | 27,108 tokens | 77.9 | 2 |
Frequently asked questions
- How fast is NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e for local inference?
- Across 2 community benchmark runs, NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e reaches up to 106.6 tok/s and averages 106.5 tok/s, with the fastest results on Apple M5 Max.
- How much memory does NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e need?
- The leanest observed configuration used about 31.8 GB of memory.
- Which tools have been used to run NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mlx-oQ8e?
- Benchmarks were submitted using oMLX. Results are community-contributed and updated as new runs arrive.