NVIDIA model product
Nemotron 3 Super is NVIDIA’s 120B total / 12B active parameter hybrid Mamba-Attention Mixture-of-Experts model for agentic reasoning, coding, planning, tool calling, and long-context analysis.
Updated Sep 8, 2026. Default version: Nemotron 3 Super (120B A12B)
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Nemotron 3 Super (120B A12B).
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkWMT24++ | Score86.67% | Rank01 | Participants23 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkRULER | Score91.75% | Rank02 | Participants4 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkArena-Hard v2 | Score73.88% | Rank06 | Participants19 | Percentile72.22% | EvidenceC | Evaluated |
| BenchmarkBird-SQL (dev) | Score41.80% | Rank06 | Participants9 | Percentile37.50% | EvidenceC | Evaluated |
| BenchmarkHMMT 2025 | Score94.73% | Rank08 | Participants33 | Percentile78.13% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench | Score81.19% | Rank09 | Participants75 | Percentile89.19% | EvidenceC | Evaluated |
| BenchmarkMMLU-ProX | Score79.36% | Rank12 | Participants32 | Percentile64.52% | EvidenceC | Evaluated |
| BenchmarkMulti-Challenge | Score55.23% | Rank12 | Participants29 | Percentile60.71% | EvidenceC | Evaluated |
| BenchmarkSciCode | Score42.05% | Rank12 | Participants24 | Percentile52.17% | EvidenceC | Evaluated |
| BenchmarkAA-LCR | Score58.31% | Rank14 | Participants19 | Percentile27.78% | EvidenceC | Evaluated |
| BenchmarkTau2 Airline | Score56.25% | Rank19 | Participants24 | Percentile21.74% | EvidenceC | Evaluated |
| BenchmarkIFBench | Score72.56% | Rank21 | Participants39 | Percentile47.37% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench | Score25.78% | Rank22 | Participants25 | Percentile12.50% | EvidenceC | Evaluated |
| BenchmarkTau2 Retail | Score62.83% | Rank25 | Participants27 | Percentile7.69% | EvidenceC | Evaluated |
| BenchmarkTau2 Telecom | Score64.36% | Rank30 | Participants36 | Percentile17.14% | EvidenceC | Evaluated |
| BenchmarkMMLU-Pro | Score83.73% | Rank32 | Participants138 | Percentile77.37% | EvidenceC | Evaluated |
| BenchmarkSWE-bench Multilingual | Score45.78% | Rank38 | Participants43 | Percentile11.90% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score90.21% | Rank48 | Participants119 | Percentile60.17% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.0 | Score31.00% | Rank51 | Participants51 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score31.28% | Rank59 | Participants63 | Percentile6.45% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam | Score22.82% | Rank62 | Participants103 | Percentile40.20% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score82.70% | Rank79 | Participants247 | Percentile68.29% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score53.73% | Rank93 | Participants113 | Percentile17.86% | EvidenceC | Evaluated |
| BenchmarkLM Arena Text Factuality | Score1,389.64 rating | Rank108 | Participants121 | Percentile10.83% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text | Score1,377.90 rating | Rank114 | Participants210 | Percentile45.93% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,360.69 rating | Rank136 | Participants210 | Percentile35.41% | EvidenceA | Evaluated |
Preference and agent-evaluation results for the default version.
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext | Categoryspanish | Rank85 | Rating / score1,415.38 | Votes347 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry entertainment and sports and media | Rank97 | Rating / score1,350.17 | Votes1,141 | ObservationsN/A | Result date |
| Arenatext factuality | Categorymulti turn | Rank97 | Rating / score1,383.79 | Votes1,263 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry business and management and financial operations | Rank99 | Rating / score1,387.77 | Votes1,348 | ObservationsN/A | Result date |
| Arenatext | Categoryfrench | Rank101 | Rating / score1,403.09 | Votes268 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycoding | Rank101 | Rating / score1,440.32 | Votes1,720 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry software and it services | Rank102 | Rating / score1,433.67 | Votes2,659 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank103 | Rating / score1,428.23 | Votes499 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryenglish | Rank103 | Rating / score1,407.17 | Votes2,988 | ObservationsN/A | Result date |
| Arenatext factuality | Categorylonger query | Rank103 | Rating / score1,390.30 | Votes2,026 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts english | Rank104 | Rating / score1,419.51 | Votes1,657 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry writing and literature and language | Rank104 | Rating / score1,351.89 | Votes1,571 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts | Rank105 | Rating / score1,410.94 | Votes4,114 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryinstruction following | Rank105 | Rating / score1,371.57 | Votes1,946 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank107 | Rating / score1,396.08 | Votes603 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry mathematical | Rank107 | Rating / score1,402.12 | Votes412 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryexclude ties | Rank107 | Rating / score1,368.05 | Votes5,116 | ObservationsN/A | Result date |
| Arenatext factuality | Categorynon english | Rank107 | Rating / score1,371.29 | Votes4,416 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank108 | Rating / score1,389.64 | Votes7,432 | ObservationsN/A | Result date |
| Arenatext | Categoryenglish | Rank109 | Rating / score1,400.02 | Votes3,032 | ObservationsN/A | Result date |
| Arenatext style control | Categoryspanish | Rank109 | Rating / score1,385.43 | Votes347 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry legal and government | Rank110 | Rating / score1,392.48 | Votes520 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry software and it services | Rank110 | Rating / score1,410.63 | Votes2,666 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank112 | Rating / score1,397.57 | Votes1,697 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank113 | Rating / score1,404.55 | Votes1,776 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank113 | Rating / score1,387.90 | Votes4,128 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry life and physical and social science | Rank113 | Rating / score1,388.44 | Votes1,134 | ObservationsN/A | Result date |
| Arenatext | Categorypolish | Rank113 | Rating / score1,348.33 | Votes267 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank114 | Rating / score1,377.90 | Votes7,458 | ObservationsN/A | Result date |
| Arenatext | Categoryexclude ties | Rank115 | Rating / score1,346.68 | Votes5,136 | ObservationsN/A | Result date |
Official vendor API pricing appears first, followed by individual provider offers.
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderKenari | Provider model IDnemotron-3-super-120b-a12b | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderRequesty | Provider model IDnemotron-3-super-120b-a12b | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderNanoGPT | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.05 | Output / 1M$0.25 | Context262.1K | Updated |
| ProviderOpenRouter | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.085 | Output / 1M$0.40 | Context1M | Updated |
| ProviderKilo Gateway | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.085 | Output / 1M$0.40 | Context262.1K | Updated |
| ProviderPioneer | Provider model IDnvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | Regionglobal | Input / 1M$0.09 | Output / 1M$0.45 | Context256K | Updated |
| ProviderVercel AI Gateway | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.15 | Output / 1M$0.65 | Context256K | Updated |
| ProviderNvidia | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.20 | Output / 1M$0.80 | Context262.1K | Updated |
| ProviderPerplexity Agent | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.25 | Output / 1M$2.5 | Context1M | Updated |
| ProviderSynthetic | Provider model IDhf:nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 | Regionglobal | Input / 1M$0.30 | Output / 1M$1 | Context262.1K | Updated |
| ProviderNebius Token Factory | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.30 | Output / 1M$0.90 | Context262.1K | Updated |
| ProviderTensorX | Provider model IDnvidia/nemotron-3-super-120b-a12b | Regionglobal | Input / 1M$0.30 | Output / 1M$0.90 | Context262.1K | Updated |
| ProviderCrusoe | Provider model IDnvidia/NVIDIA-Nemotron-3-Super-120B-A12B | Regionglobal | Input / 1M$0.30 | Output / 1M$2.4 | Context262.1K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Provider-specific output speed and catalog latency for Nemotron 3 Super (120B A12B). Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Technical details for the model's default version.
Available versions of this model. The score column identifies the version used in the overall ranking.
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionNemotron 3 Super (120B A12B) | Released | LLMBoard44.28 | Parameters120B | Context262.1K | Max output262.1K | Open weightsYes | LicenseNVIDIA Open Model License Agreement |
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Key information about Nemotron 3 Super and its available data.
Nemotron 3 Super is a hybrid Mamba-Attention Mixture-of-Experts model with Mamba-2 layers, Transformer attention layers, LatentMoE routing, Multi-Token Prediction, and native NVFP4 pretraining on Blackwell.
It supports a 1M-token context window and was pre-trained on 25 trillion tokens and post-trained with multi-environment reinforcement learning across 21 configurations.
Data as of 2026-09-08.
Common questions about Nemotron 3 Super.
Nemotron 3 Super's default version was released on Mar 11, 2026.
Nemotron 3 Super's official API price is $0.20 per million input tokens and $0.80 per million output tokens via Nvidia. The lowest tracked third-party offer starts at $0.05 input and $0.25 output via NanoGPT.
Nemotron 3 Super was created by NVIDIA.
The default version has a 262.1K token context window.
Yes. The default version is marked as open weight under NVIDIA Open Model License Agreement.
13 provider offerings are linked to the default version.