Blog

NVIDIA Nemotron 3 Super Is Now Available in Puter.js

On this page

Puter.js now supports Nemotron 3 Super, NVIDIA's latest open model built for complex agentic AI systems.

What is NVIDIA Nemotron 3 Super?

Nemotron 3 Super is a 120-billion-parameter open model with only 12 billion active parameters, designed for running complex multi-agent systems at scale. It uses a hybrid Latent Mixture-of-Experts (LatentMoE) architecture that interleaves Mamba-2 and MoE layers to deliver up to 5x higher throughput than the previous Nemotron Super. With a native 1-million-token context window, agents get long-term memory for aligned, high-accuracy reasoning without goal drift.

The model is pretrained on 25 trillion tokens using NVFP4, NVIDIA's 4-bit floating-point format optimized for the Blackwell architecture. It excels at reasoning, coding, tool use, and cybersecurity triaging tasks.

Examples

Basic Chat

puter.ai.chat("Explain the concept of machine learning in simple terms", {
    model: "nvidia/nemotron-3-super-120b-a12b:free"
})

Agentic Reasoning

puter.ai.chat("Break down the steps to debug a memory leak in a Node.js application and suggest tools to use",
    { model: "nvidia/nemotron-3-super-120b-a12b:free" }
)

Code Generation

puter.ai.chat("Write a Python function that implements binary search with error handling",
    { model: "nvidia/nemotron-3-super-120b-a12b:free" }
)

Get Started Now

Just add one library to your project:

// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';

Or add one script tag to your HTML:

<script src="https://js.puter.com/v2/"></script>

No API keys or NVIDIA account needed. Start building with Nemotron 3 Super immediately.

Learn more:

Free, Serverless AI and Cloud

Start creating powerful web applications with Puter.js in seconds!

Get Started Now

Read the Docs Try the Playground