Skip to content

NVIDIA · Chat

Nemotron Super 3 120b

NVIDIA 🇺🇸ChatOpen weights. One of its sources says the weights are publicly available.
From Neosantara's listing:
Nemotron 3 Super 120B is a large-scale MoE reasoning model by NVIDIA for complex reasoning, programming, and mathematics.
Market median price
$0.33 blended / 1M
Max context
1M
Listings
7
Providers
5

Evidence and sources

Maker HQ: US (source)
Sources: AWS Bedrock Runtime data, FlowDot data, Models.dev data, Neosantara Inference dataPrimary data from AWS Bedrock Runtime from FlowDot: name from Models.dev: weights availability from Neosantara Inference: description

Advantages and trade-offs

Compared with 10 representative chat models

Advantages

Open weights available

The weights are public. 4 of the 10 peers that say either way have open weights.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 6 maker's prices; 1 market median price.

Input & output

as stated by the model's providers
Text
Text

Capabilities

Supported operations and provider-stated features

ReasoningFunction callingStructured outputBatch API

Listings

 Amazon Web Servicesnvidia.nemotron-super-3-120bSources: AWS Bedrock Runtime data, LiteLLM data, Models.dev data256K$0.15batch $0.075$0.65batch $0.325
 Amazon Web Servicesnvidia.nemotron-super-3-120b:0Source: GenAI Prices dataNot published by the provider$0.15$0.65
 Amazon Web ServicesRegions: 🇺🇸 us-govus-gov.nvidia.nemotron-super-3-120bSource: LiteLLM data256K$0.18$0.78
 CrusoeNemotron 3 Super 120BSource: Crusoe Managed InferenceNot published by the provider$0.30cache read $0.15$2.40
 Neosantaranemotron-3-super-120bSource: Neosantara Inference data128K
 T SystemsNemotron-3-Super-120B-FP8 · fp8Source: T-Systems LLM Hub1M

1 listing without a price is hidden.

These providers advertise this model. Researchers have found some resellers serving a different, cheaper model, so see how to check whether a provider meets your needs.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Science

Data credits: ChemBench leaderboard, Jablonka group / lamalab-org. Public verified numeric result summaries; results licence not separately stated.

Around the web

  • nemotron-3-super-120b-a12b Model by NVIDIA | NVIDIA NIM
    nemotron-3-super-120b-a12b Model by NVIDIA · NVIDIA NIM Open, efficient hybrid Mamba-Transformer MoE with 1M context, …nemotron-3-super-120b-a12b Model by NVIDIA · NVIDIA NIM Open, efficient hybrid Mamba-Transformer MoE with 1M context, excelling in agentic reasoning, coding, planning, tool calling, and more

    build.nvidia.com

  • nemotron-3-super-120b-a12b
    title: "nemotron-3-super-120b-a12b" publisher: "nvidia" type: "endpoint" updated: "2026-03-11T15:40:20.019Z" description: "Open, efficient hybrid Mamba-Transformer MoE with …title: "nemotron-3-super-120b-a12b" publisher: "nvidia" type: "endpoint" updated: "2026-03-11T15:40:20.019Z" description: "Open, efficient hybrid Mamba-Transformer MoE with 1M context, excelling in a…

    build.nvidia.com

  • nemotron-3-super-120b-a12b Model by NVIDIA | NVIDIA NIM
    title: "nemotron-3-super-120b-a12b" publisher: "nvidia" type: "endpoint" updated: "2026-03-11T15:40:20.019Z" description: "Open, efficient hybrid Mamba-Transformer MoE with …title: "nemotron-3-super-120b-a12b" publisher: "nvidia" type: "endpoint" updated: "2026-03-11T15:40:20.019Z" description: "Open, efficient hybrid Mamba-Transformer MoE with 1M context, excelling in a…

    build.nvidia.com

  • nemotron-3-super-120b-a12b Model by NVIDIA | NVIDIA NIM
    nemotron-3-super-120b-a12b Model by NVIDIA · NVIDIA NIM Description Nemotron-3-Super-120B-A12B-BF16 is a large language model (LLM) …nemotron-3-super-120b-a12b Model by NVIDIA · NVIDIA NIM Description Nemotron-3-Super-120B-A12B-BF16 is a large language model (LLM) trained by NVIDIA, designed to deliver strong agentic, reasoning, a…

    build.nvidia.com

  • nvidia / nemotron-3-super-120b-a12b
    nemotron-3-super-120b-a12b NVIDIA-Nemotron-3-Super-120B-A12B Model Summary · Total Parameters · 120B (12B active) Architecture · LatentMoE - …nemotron-3-super-120b-a12b NVIDIA-Nemotron-3-Super-120B-A12B Model Summary · Total Parameters · 120B (12B active) Architecture · LatentMoE - Mamba-2 + MoE + Attention hybrid with Multi-Token Predicti…

    docs.api.nvidia.com

  • Get Started with Nemotron 3 Super 120B — NVIDIA NIM for Large Language Models
    d '{ "model": "nvidia/nemotron-3-super-120b-a12b", "messages": [ { "role":"user", "content":"Write a haiku about GPUs" } ], …d '{ "model": "nvidia/nemotron-3-super-120b-a12b", "messages": [ { "role":"user", "content":"Write a haiku about GPUs" } ], "temperature": 1.0, "top_p": 0.95, "max_tokens": 1024, "chat_template_kwarg…

    docs.nvidia.com

  • Nemotron-3-Super-120B-A12B (Experimental) | NVIDIA NGC
    License/Terms of Use: GOVERNING TERMS: The NIM container is governed by the NVIDIA Software License …License/Terms of Use: GOVERNING TERMS: The NIM container is governed by the NVIDIA Software License Agreement and Product-Specific Terms for NVIDIA AI Products . Use of this model is governed by the …

    catalog.ngc.nvidia.com

  • Using Nemotron-3-Super-120B-A12B LLM NIM — NVIDIA RAG blueprint
    services using self-hosted on-premises models # Update nims.yaml Edit deploy/compose/nims.yaml and change the nim-llm service …services using self-hosted on-premises models # Update nims.yaml Edit deploy/compose/nims.yaml and change the nim-llm service image and GPU allocation: nim-llm : image : nvcr.io/nim/nvidia/nemotron-3…

    docs.nvidia.com

  • Nemotron-3-Super-120B-A12B - NGC Catalog - NVIDIA
    NGC Catalog Explore Search Support API Catalog Forum Welcome Guest NVIDIA NVIDIA Nemotron-3-Super-120B-A12B Get Container …NGC Catalog Explore Search Support API Catalog Forum Welcome Guest NVIDIA NVIDIA Nemotron-3-Super-120B-A12B Get Container Subscription required to access this content Container NVIDIA NVIDIA Nemotron…

    catalog.ngc.nvidia.com

  • Nemotron 3 Super 120B-A12B | Jetson AI Lab
    60 GB of weights plus KV cache require the 128 GB SKU Nemotron 3 Family …60 GB of weights plus KV cache require the 128 GB SKU Nemotron 3 Family Model · Parameters · Memory · Best For Nemotron3 Nano 4B · 4B · 4GB RAM · Lightweight edge deployment Nemotron3 Nano 30B-A3B · …

    jetson-ai-lab.com

Also known as: bedrock/ap-northeast-1/nvidia.nemotron-super-3-120b, bedrock/ap-south-1/nvidia.nemotron-super-3-120b, bedrock/ap-southeast-2/nvidia.nemotron-super-3-120b, bedrock/ap-southeast-3/nvidia.nemotron-super-3-120b, bedrock/eu-central-1/nvidia.nemotron-super-3-120b, bedrock/eu-north-1/nvidia.nemotron-super-3-120b, bedrock/eu-south-1/nvidia.nemotron-super-3-120b, bedrock/eu-west-1/nvidia.nemotron-super-3-120b, bedrock/eu-west-2/nvidia.nemotron-super-3-120b, bedrock/sa-east-1/nvidia.nemotron-super-3-120b, bedrock/us-gov-east-1/nvidia.nemotron-super-3-120b, bedrock/us-gov-west-1/nvidia.nemotron-super-3-120b, bedrock_mantle/ap-northeast-1/nvidia.nemotron-super-3-120b, bedrock_mantle/ap-south-1/nvidia.nemotron-super-3-120b, bedrock_mantle/ap-southeast-2/nvidia.nemotron-super-3-120b, bedrock_mantle/ap-southeast-3/nvidia.nemotron-super-3-120b, bedrock_mantle/eu-central-1/nvidia.nemotron-super-3-120b, bedrock_mantle/eu-north-1/nvidia.nemotron-super-3-120b, bedrock_mantle/eu-south-1/nvidia.nemotron-super-3-120b, bedrock_mantle/eu-west-1/nvidia.nemotron-super-3-120b, bedrock_mantle/eu-west-2/nvidia.nemotron-super-3-120b, bedrock_mantle/nvidia.nemotron-super-3-120b, bedrock_mantle/sa-east-1/nvidia.nemotron-super-3-120b, bedrock_mantle/us-gov-east-1/nvidia.nemotron-super-3-120b, bedrock_mantle/us-gov-west-1/nvidia.nemotron-super-3-120b, nemotron-3-super-120b, nemotron-3-super-120b-fp8, nvidia.nemotron-super-3-120b, nvidia.nemotron-super-3-120b:0, nvidia/nemotron-super-3-120b, us-gov.nvidia.nemotron-super-3-120b

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works