Nemotron 3 Super
Editorial notes
Lanzado 11 marzo 2026 en GTC 2026. Hybrid Mamba-Transformer MoE 120B / 12B activos (es el Super 120B-A12B). LatentMoE architecture. 5x throughput NVFP4 Blackwell. Supera GPT-OSS-120B con +10% throughput/GPU. #1 DeepResearch Bench. Scores del model card oficial nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16 (AIME25 90.2, LiveCodeBench v5 81.2, HLE sin tools 18.3, SWE-V scaffold OpenHands 60.5). Auditoria 2026-06-08: MMLU-Pro corregido 83.3->83.7 y GPQA-D 79.4->79.2 (transcripcion imprecisa vs el 83.73/79.23 del card).
Spec sheet
- Company
- Nvidia
- Country
- US
- Type
- reasoning
- Release
- 2026-03
- Context
- 1.0M tokens
- Params total
- 120B
- Params active (MoE)
- 12B
- License
- nvidia-open-model
- Quants
- BF16, Q8_0, Q5_K_M, Q4_K_M
- Pricing (openrouter)
- $0.09/$0.45/M
- Slug
- nemotron-3-super
Quick install
2 toolsollama run nemotron-3:superNote: ~70 GB VRAM Q4_K_M
vllm serve nvidia/Nemotron-3-Super --tensor-parallel-size 2Benchmarks (7)
Reasoning 3
Coding 2
Cite this model
BibTeX · APA
BibTeX
@misc{frontier-nemotron-3-super,
title = {Nemotron 3 Super},
author = {{Nvidia}},
year = {2026},
note = {Frontier Benchmarks AI atlas. Accessed 2026-06-17},
url = {https://frontierbenchmarks.com/models/nemotron-3-super}
} APA
Nvidia (2026). Nemotron 3 Super [Large language model]. Frontier Benchmarks AI. Retrieved 2026-06-17, from https://frontierbenchmarks.com/models/nemotron-3-super
Citation reflects the atlas page, not the original model paper. For the paper, see the "Resources" section above.