n8n-nodes-nemotron

This community node integrates NVIDIA's powerful Nemotron Ultra 253B model with n8n, the workflow automation platform. Access state-of-the-art AI capabilities directly in your workflows and agents.

Overview

This package adds the NVIDIA Nemotron Ultra 253B model to n8n's AI toolkit, allowing you to:

• Use one of the most powerful open-source AI models in your n8n workflows
• Connect Nemotron Ultra to n8n's AI Agents and AI Chains
• Optimize conversations with fine-tuned parameters
• Process natural language within your automation workflows

Installation

Follow these steps to install this custom node:

In n8n v1.0+:

1. Go to Settings > Community Nodes
2. Click on Install a community node
3. Enter n8n-nodes-nemotron in the "Install npm package" input field
4. Click on Install

Manual installation:

# Navigate to your n8n installation folder
cd ~/.n8n

# Install the package
npm install n8n-nodes-nemotron

Configuration

To use the Nemotron Ultra model, you need to set up your NVIDIA API credentials:

1. Go to your n8n instance
2. Open the Credentials page
3. Click on Create New Credentials
4. Search for "NVIDIA API" and select it
5. Enter your NVIDIA API Key
6. Set the Base URL to https://integrate.api.nvidia.com/v1
7. Click on Save

Usage

The NVIDIA Nemotron node can be used in two primary ways:

1. With AI Agent

1. Add the Bro AI Agent node to your workflow
2. Add the NVIDIA Nemotron node to your workflow
3. Connect the Nemotron node to the Agent node
4. Configure the Agent as needed
5. Run your workflow

2. With AI Chain

1. Add the AI Chain node to your workflow
2. Add the NVIDIA Nemotron node to your workflow
3. Connect the Nemotron node to the Chain node
4. Configure your prompt and other settings
5. Run your workflow

Available Options

The Nemotron Ultra model can be configured with these parameters:

• Temperature - Controls response randomness (default: 0.6)
• Top P - Controls token diversity via nucleus sampling (default: 0.95)
• Maximum Tokens - Limits the length of generated responses (default: 4096)
• Frequency Penalty - Reduces repetition in responses (default: 0)
• Presence Penalty - Encourages the model to talk about new topics (default: 0)
• System Message - Sets the behavior/role of the assistant
• Response Format - Select between regular text or structured JSON output

Example Workflow

Here's a simple example workflow using the Nemotron node with an AI Agent:

[Manual Trigger] → [Bro AI Agent] ← [NVIDIA Nemotron]
                     ↓
                [Set] → [Respond to Webhook]

Compatibility

• Requires n8n version 1.0.0 or later
• Works with Bro AI Agent and standard n8n AI Chain nodes
• Compatible with browsers supporting modern JavaScript features

Contributing

Contributions are welcome! If you encounter any issues or have suggestions for improvements:

1. Check the GitHub repository
2. Submit an issue describing your problem or suggestion
3. Create a pull request if you have a solution to propose

License

This project is licensed under the MIT License - see the LICENSE file for details.

Credits

• Developed by Bro Lendario
• Email: [email protected]
• Powered by NVIDIA's Nemotron Ultra 253B LLM technology
• Built for n8n workflow automation platform

FAQ

Q: Do I need a NVIDIA account to use this node?
A: Yes, you need to register for NVIDIA's API services and get an API key.

Q: Is the NVIDIA Nemotron Ultra model free to use?
A: You'll need to check NVIDIA's current pricing and usage terms.

Q: Can I use this with multiple LLMs in the same workflow?
A: Yes, you can use multiple LLM nodes in an n8n workflow and choose which one to connect to your agents or chains.

Q: Does this support streaming responses?
A: Yes, streaming is enabled by default for a better user experience.

If you find this node useful, please consider starring the repository and sharing it with others who might benefit from using NVIDIA's Nemotron Ultra model with n8n.