DeepSeek & Llama

Intellinode supports NVIDIA’s latest language models, Deepseek and Llama, via a unified chatbot interface. With minimal code changes, you can switch between NVIDIA, OpenAI, and other providers.

Supported Models

Sample of supported models with much more available using Intellinode Nvidia connector:

Model Name
deepseek-ai/deepseek-v4-flash-0731
meta/llama-3.3-70b-instruct
tiiuae/falcon3-7b-instruct

Get Started

API Key

Visit NVIDIA model catalog to get your API key.

Chat Code

const { Chatbot, NvidiaInput, SupportedChatModels } = require("intellinode");

Provide your NVIDIA API key and create a chatbot instance:

const nvidiaBot = new Chatbot(NVIDIA_API_KEY, SupportedChatModels.NVIDIA);

Construct a chat input using the NvidiaInput class and add your message(s):

const input = new NvidiaInput("You are a helpful assistant.", {
  model: 'deepseek-ai/deepseek-v4-flash-0731', // Use Deepseek or NVIDIA Llama model
  maxTokens: 512,
  temperature: 0.6
});
input.addUserMessage("Which number is larger, 9.11 or 9.8?");

Send the chat input:

const response = await nvidiaBot.chat(input);
console.log(response); // Returns a plain text response with any <think> tags removed

Multiple Messages

Nvidia Chat supports multi-turn conversations just like other chatbot models:

const input = new NvidiaInput("You are an insightful assistant.", {
  model: 'deepseek-ai/deepseek-v4-flash-0731',
  maxTokens: 512,
  temperature: 0.6
});
input.addUserMessage("What's the summary of the Inception movie?");
input.addAssistantMessage("Inception is about a thief who enters dreams to extract or plant ideas.");
input.addUserMessage("How does that compare to Interstellar?");
const responses = await nvidiaBot.chat(input);
responses.forEach(resp => console.log("- " + resp));

Try DeepSeek From the UI

To chat with DeepSeek models without code, open IntelliChat, select the DeepSeek provider and paste your DeepSeek key. See Chat UI for the setup.

NVIDIA NIM

Nvidia NIM provide optimized way to host models locally. Download NVIDIA NIM as instructed in Nvidia documentation.

Update your client to point to your local endpoint:

const { Chatbot, NvidiaInput, SupportedChatModels } = require("intellinode");

// provide your NVIDIA key and local host url
const nvidiaBot = new Chatbot(NVIDIA_API_KEY, SupportedChatModels.NVIDIA, null, { baseUrl: 'http://0.0.0.0:8000'});

// construct a chat input using the `NvidiaInput`
const input = new NvidiaInput("You are a helpful assistant.", {
  model: 'meta/llama-3.3-70b-instruct'
});
input.addUserMessage("Which number is larger, 9.11 or 9.8?");

// call the chatbot
const response = await nvidiaBot.chat(input);

Check available NIM models in NVIDIA's model catalog. Open any model and follow the setup instructions under the Docker tab to deploy.