Nemotron 3.5 Lightning: benchmark results
NVIDIA's Nemotron 3.5 Lightning, a 30B-A3B hybrid Mamba-2/MoE open model released in full BF16 precision as the reference checkpoint for further fine-tuning, distillation or quantization. Provider: NVIDIA. Released 2026-08-11. Access: Open.
Aggregate Rating 1594 ± 15, rank #455 of 1635 rated models, from 130 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Hallucination Rate | 37.57 | Hallucination Rate (%) | 88.3 |
| SpeechMap Compliance | 75 | % Requests Completed | 71 |
| MLCR-AA - Conciseness | 97.22 | Conciseness (%) | 67.9 |
| AA-Omniscience Index - Business | -12.5 | Omniscience Index | 66.6 |
| AA-Omniscience Index - Law | -16 | Omniscience Index | 65.4 |
| AA-Omniscience Index - Software Engineering (SWE) - Dart | -12 | Omniscience Index | 64.8 |
| AA-Omniscience Index - Humanities & Social Sciences | -12.7 | Omniscience Index | 64 |
| AI for Education Pedagogy - Social studies | 81.82 | Accuracy (%) | 62.7 |
| AA-Omniscience Index - Science, Engineering & Mathematics | -9.8 | Omniscience Index | 62.5 |
| AA Omniscience | -17.72 | Score | 59.6 |
| AA-Omniscience Index - Health | -26.7 | Omniscience Index | 58.3 |
| AI Chess Leaderboard (Reasoning) | 752 | Elo | 57.5 |
Interactive version: theaggregate.ai/model?slug=nemotron-3-5-lightning · How It Works · Data refreshed daily, snapshot 2026-10-15.