Llama 3.3 Nemotron Super 49B (v1) (Non-reasoning): benchmark results
Provider: NVIDIA. Released 2025-03-18. Access: Open.
Aggregate Rating 1438 ± 18, rank #1320 of 2201 rated models, from 44 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| NVIDIA Nemotron 49B Report - BFCL V2 Live | 73.9 | Score (%) | 100 |
| AA-Omniscience Hallucination Rate | 73.69 | Hallucination Rate (%) | 57 |
| NVIDIA Nemotron 49B Report - IFEval | 89 | Score (%) | 50 |
| AA Omniscience - Software Engineering (SWE) - PHP | 20 | Accuracy (%) | 49.5 |
| AA Omniscience - Software Engineering (SWE) - HTML | 28 | Accuracy (%) | 42.7 |
| AA Omniscience - Software Engineering (SWE) - C | 28 | Accuracy (%) | 41.8 |
| AA Omniscience - Software Engineering (SWE) - Java | 13 | Accuracy (%) | 38.9 |
| AA Omniscience - Software Engineering (SWE) - Python | 14.5 | Accuracy (%) | 37.4 |
| AA IFBench | 39.46 | Accuracy (%) | 37.1 |
| AA MMLU-Pro | 69.77 | Accuracy (%) | 36.2 |
| AA Omniscience - Software Engineering (SWE) - Go | 12 | Accuracy (%) | 34.5 |
| AA MATH-500 | 77.53 | Accuracy (%) | 33.3 |
Interactive version: theaggregate.ai/model?slug=llama-3-3-nemotron-super-49b-v1-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-12.