Nemotron 3 Ultra 550B: benchmark results
Provider: NVIDIA.
Aggregate Rating 1717 ± 38, rank #279 of 10916 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Wolfram LLM Benchmarking Project | 61.8 | Correct Functionality (%) | 84.7 |
| THOR Finding Triage - False Review Load | 37.1 | False positives not suppressed (%, lower is better) | 74.4 |
| THOR Finding Triage - CW% | 59.5 | Confidence-weighted classification score (%) | 69.3 |
| THOR Finding Triage - Balanced OTS | 60.3 | Class-balanced operational triage score (%) | 61.4 |
| TradeVerse - HS Code Prediction | 63.21 | Micro-F1 (%; multi-label prediction of the Harmonized System | 60 |
| TradeVerse - Statement Generation | 57.83 | BERTScore F1 (%; microsoft/deberta-large-mnli, no baseline r | 60 |
| THOR Finding Triage - Critical Miss Rate | 1.8 | True positives suppressed as false positives (%, lower is be | 42 |
| TradeVerse - Respondent Identification (All Members Masked) | 90.6 | Accuracy (%; naming the responding WTO member from the anony | 40 |
| TradeVerse - Respondent Identification (Respondent Masked) | 90.68 | Accuracy (%; naming the responding WTO member from the anony | 40 |
| React Native Evals | 61.43 | Overall Score (%) | 6.5 |
Interactive version: theaggregate.ai/model?slug=nemotron-3-ultra-550b · How It Works · Data refreshed daily, snapshot 2026-11-08.