Nemotron 3 Nano Omni 30B-A3B (Reasoning) — benchmark results
Provider: NVIDIA. Released 2026-04-28. Access: Open.
Unified ELO 1528 ± 22, rank #691 of 1776 rated models, from 36 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA IFBench | 63.2 | Accuracy (%) | 74.3 |
| AA Omniscience - Software Engineering (SWE) - Kotlin | 18 | Accuracy (%) | 50 |
| Artificial Analysis Intelligence Index | 14.93 | Intelligence Index | 49.4 |
| AA TAU-2 Bench | 45.32 | Accuracy (%) | 49.1 |
| AA Long Context Reasoning | 35.67 | Accuracy (%) | 47 |
| AA Terminal-Bench Hard | 8.33 | Accuracy (%) | 41.8 |
| AA Omniscience - Software Engineering (SWE) - Rust | 46 | Accuracy (%) | 41.5 |
| AA Omniscience - Software Engineering (SWE) - Java | 15 | Accuracy (%) | 41.3 |
| AA Humanity's Last Exam | 5.33 | Accuracy (%) | 39.3 |
| AA Omniscience - Software Engineering (SWE) - R | 8 | Accuracy (%) | 37.9 |
| AA Omniscience - Science, Engineering & Mathematics | 22.5 | Accuracy (%) | 36.5 |
| AA SciCode | 27.78 | Accuracy (%) | 36.4 |
Interactive version: aibenchmarks.dev/model?slug=nemotron-3-nano-omni-30b-a3b-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.