nemotron-3-nano-30b-a3b

Approved Data Classifications

Description

Nemotron-3-Nano-30B-A3B-BF16 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 configured through a flag in the chat template. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so, albeit with a slight decrease in accuracy for harder prompts that require reasoning. Conversely, allowing the model to generate reasoning traces first generally results in higher-quality final solutions to queries and tasks.

The model employs a hybrid Mixture-of-Experts (MoE) architecture, consisting of 23 Mamba-2 and MoE layers, along with 6 Attention layers. Each MoE layer includes 128 experts plus 1 shared expert, with 6 experts activated per token. The model has 3.5B active parameters and 30B parameters in total.

The above was sourced from the following model card on hugging face: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16

Capabilities

Model	Release Date	Input	Output	Context Length	Cost (per 1 million tokens)
nemotron-3-nano-30b-a3b	Dec 2024	Text	Text	131,072	$0.06/1M input $0.24/1M output

info

• 1M represents 1 Million Tokens
• All prices listed are based on 1 Million Tokens

Availability

Cloud Provider

Usage

curl

curl -X POST https://api.ai.it.ufl.edu/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer <API_TOKEN>" \
-d '{
    "model": "nemotron-3-nano-30b-a3b",
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant."
        },
        {
            "role": "user",
            "content": "Write a haiku about an Alligator."
        }
    ]
}'

python

from openai import OpenAI
  client = OpenAI(
      api_key="your_api_key",
      base_url="https://api.ai.it.ufl.edu/v1"
  )

  response = client.chat.completions.create(
      model="nemotron-3-nano-30b-a3b", # model to send to the proxy
      messages = [
          { role: "system", content: "You are a helpful assistant." },
          {
              "role": "user",
              "content": "Write a haiku about an Alligator."
          }
      ]
  )

  print(response.choices[0].message)

javascript

import OpenAI from 'openai';

const openai = new OpenAI({
  apiKey: 'your_api_key',
  baseURL: 'https://api.ai.it.ufl.edu/v1'
});

const completion = await openai.chat.completions.create({
    model: "nemotron-3-nano-30b-a3b",
    messages: [
        { role: "system", content: "You are a helpful assistant." },
        {
            role: "user",
            content: "Write a haiku about an Alligator.",
        },
    ],
});

print(completion.choices[0].message)