Nemotron 3 Super 120B A12B (Reasoning): benchmark results

Provider: NVIDIA. Released 2026-03-11.

Aggregate Rating 1586 ± 14, rank #800 of 2245 rated models, from 86 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
K2 Horizon MoVA Card - CritPt3.1Score (%)91.7
AA Omniscience - Software Engineering (SWE) - Julia32Accuracy (%)90.1
AA IFBench71.5Accuracy (%)87.9
AA Omniscience - Software Engineering (SWE) - TypeScript37.78Accuracy (%)84
K2 Horizon MoVA Card - AA-Omniscience Accuracy24.3Score (%)83.3
AA Omniscience - Software Engineering (SWE) - Swift48Accuracy (%)82.4
AA Omniscience - Software Engineering (SWE) - JavaScript40Accuracy (%)81.6
AA Omniscience - Software Engineering (SWE) - HTML44Accuracy (%)80.7
AA Omniscience - Software Engineering (SWE) - Kotlin28Accuracy (%)80.5
MT-JailBench - XTeaming37.74Attack success rate (%) of the XTeaming multi-turn jailbreak80
AA Omniscience - Software Engineering (SWE) - PHP34Accuracy (%)79.4
AA Omniscience - Software Engineering (SWE) - Dart26Accuracy (%)77.6

Interactive version: theaggregate.ai/model?slug=nemotron-3-super-120b-a12b-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-16.