NVIDIA modelsChat Completions model

nvidia/nemotron-3-super-120b-a12b

Name: NVIDIA: Nemotron 3 Super
Price: 0.09 USD
Availability: InStock
Author: NVIDIA

NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. Built on a hybrid Mamba-Transformer Mixtu...

Input⌘ / Ctrl + Enter

Try a prompt

Prompt

1M context262.1K max output

Input $0.09/M · Output $0.45/M

OutputReady

Your response will appear here

Choose an example or write a prompt, then click Run.

API details and access

Model Details

ProviderNVIDIA

TypeLlm

Model IDnvidia/nemotron-3-super-120b-a12b

Capabilities

InputText

OutputText

Context1,000,000

Max Output262,144

VisionNot supported

Function CallingSupported

Access details

Chat Completions

Base URLhttps://api.sandbase.ai

API Endpoint/v1/chat/completions

{
  "model": "nvidia/nemotron-3-super-120b-a12b",
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ],
  "max_tokens": 512
}

Pricing

Input$0.09 / 1M tokens

Output$0.45 / 1M tokens

Production endpoint

Send your first request

OpenAI-compatible endpoint with unified authentication and usage tracking.

POSThttps://api.sandbase.ai/v1/chat/completions

Model IDnvidia/nemotron-3-super-120b-a12b

Input Schema

13 parameters · 1 required · 12 optional

Parameter	Type	Required	Description
`messages`	object[]	Required	—
`seed`	integer	Optional	—
`stop`	string[]	Optional	—
`tools`	array	Optional	—
`top_k`	integer	Optional	Min: 0
`top_p`	number	Optional	Min: 0 · Max: 1
`stream`	boolean	Optional	Default: false
`max_tokens`	integer	Optional	Max: 262144
`temperature`	number	Optional	Min: 0 · Max: 2 · Default: 1
`tool_choice`	string	Optional	—
`response_format`	object	Optional	—
`presence_penalty`	number	Optional	Min: -2 · Max: 2
`frequency_penalty`	number	Optional	Min: -2 · Max: 2

Output Schema

Field	Type	Description
`id`	string	Unique identifier for the completion
`object`	string	Object type, e.g. "chat.completion"
`model`	string	Model used for the completion
`choices`	array	List of completion choices
`choices[].message.content`	string	Generated text content
`choices[].finish_reason`	string	Reason the generation stopped
`usage.prompt_tokens`	integer	Number of tokens in the prompt
`usage.completion_tokens`	integer	Number of tokens in the completion
`usage.total_tokens`	integer	Total tokens used

Code Examples

Ready-to-run snippets

curl https://api.sandbase.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "nvidia/nemotron-3-super-120b-a12b",
    "messages": [{"role": "user", "content": "Hello"}],
    "temperature": 0.7,
    "max_tokens": 2048,
    "stream": true
  }'

← Back to NVIDIA: Nemotron 3 Super

Related Models

nvidia/llama-3.3-nemotron-super-49b-v1.5Llama-3.3-Nemotron-Super-49B-v1.5 is a 49B-parameter, English-centric reasoning/chat model derived from Meta’s Llama-3.3-70B-Instruct with a 128K context. It’s post-trained for agentic workflows (RAG, tool calling) via SFT across math, code, science, and multi-turn chat, followed by multiple RL stages; Reward-aware Preference Optimization (RPO) for alignment, RL with Verifiable Rewards (RLVR) for step-wise reasoning, and iterative DPO to refine tool-use behavior. A distillation-driven Neural Architecture Search (“Puzzle”) replaces some attention blocks and varies FFN widths to shrink memory footprint and improve throughput, enabling single-GPU (H100/H200) deployment while preserving instruction following and CoT quality. In internal evaluations (NeMo-Skills, up to 16 runs, temp = 0.6, top_p = 0.95), the model reports strong reasoning/coding results, e.g., MATH500 pass@1 = 97.4, AIME-2024 = 87.5, AIME-2025 = 82.71, GPQA = 71.97, LiveCodeBench (24.10–25.02) = 73.58, and MMLU-Pro (CoT) = 79.53. The model targets practical inference efficiency (high tokens/s, reduced VRAM) with Transformers/vLLM support and explicit “reasoning on/off” modes (chat-first defaults, greedy recommended when disabled). Suitable for building agents, assistants, and long-context retrieval systems where balanced accuracy-to-cost and reliable tool use matter. nvidia/nemotron-3-nano-30b-a3bNVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully open with open-weights, datasets and recipes so developers can easily customize, optimize, and deploy the model on their infrastructure for maximum privacy and security.nvidia/nemotron-3-ultra-550b-a55bNVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it supports text input and output with a context window of up to 1M tokens. It is suited for long-running agentic workflows, including agent orchestration, coding agents, deep research, and complex enterprise tasks. It is particularly strong at multi-step reasoning and planning, with high-throughput inference designed for high-volume agent pipelines. It is part of the NVIDIA Nemotron family of open models for agentic AI.nvidia/nemotron-nano-9b-v2NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be controlled via a system prompt. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so.