Getting started with NVIDIA ChatRTX: a beginner's guide

Learn how to run your own private AI chatbot on an NVIDIA‑powered PC with ChatRTX – simple steps for everyday use.

4 min read·6 steps·25 June 2026

Hook: By the end of this guide you’ll have a private AI chat assistant humming on your own computer, ready to answer questions, draft emails, or brainstorm ideas. It’s perfect for anyone who’s curious about AI but hasn’t written a line of code before.

✅ Before you start

A PC with an NVIDIA RTX graphics card (the RTX series adds hardware that speeds up AI work).
An internet connection to download the software.
A free NVIDIA account – you’ll need one to grab the installation files.

Create a NVIDIA account

Visit nvidia.com and click the Sign Up button at the top right. Fill in your name, email and a password. This account acts like a library card for NVIDIA’s AI tools.

💬 Try typing this**Example:** ➤

Name: Jane Doe
Email: jane.doe@example.com
Password: (choose a strong one)

After confirming the email you’re ready to download.

Download the ChatRTX server package

Log into your NVIDIA account, go to the AI Tools section, and look for ChatRTX. Click the Download button; the file will be a compressed folder (e.g., chatrtx_server.zip).

💬 Try typing this**Example:** ➤

If you see a button that says Download ChatRTX Server, click it. The file should land in your default Downloads folder.

Install the server

Extract the zip file to a folder like C:\ChatRTX.
Inside that folder, double‑click the file named install.bat (Windows) or run install.sh (Mac/Linux) – this launches a simple installer that checks your GPU and sets up the required libraries.

💬 Try typing this**Example:** ➤

C:\> cd C:\ChatRTX
C:\ChatRTX> install.bat

You’ll see messages confirming that the GPU driver (software that makes the GPU talk to the computer) and the runtime (the engine that runs the AI model) are ready.

Pull the ChatRTX model

A model is the brain of the chatbot – for ChatRTX it’s a large language model (LLM), which works like the engine behind ChatGPT. In the command prompt, type:

chatrtx pull --model default

The tool will download the model files (about a few gigabytes) onto your PC. This only happens once.

💬 Try typing this**Example:** ➤

> chatrtx pull --model default
Downloading… 2.3 GB [=========>] 100%

Start the chatbot locally

Now launch the server with a single command:

chatrtx run --port 8080

The server will listen on port 8080 (a door on your computer) and you’ll see a line that says “ChatRTX ready at http://localhost:8080”. localhost means “this computer”.

💬 Try typing this**Example:** ➤

> chatrtx run --port 8080
ChatRTX is now running at http://localhost:8080

Talk to your new assistant

Open any web browser (Chrome, Edge, Firefox) and navigate to http://localhost:8080. You’ll see a simple chat window. Type a question like “What’s a quick recipe for banana bread?” and hit Enter. The AI will reply directly in the window.

💬 Try typing this**Example:** ➤

You: What’s a quick recipe for banana bread?
ChatRTX: Here’s a simple 3‑ingredient recipe…

⚠️ Common mistakes

Forgetting to run the installer as an administrator. The installer needs permission to add drivers; right‑click install.bat and choose Run as administrator if you see an error.
Using the wrong port. If another program already uses 8080, change the command to --port 8090 and open that address in the browser instead.
Skipping the model download. Without the LLM the server starts but can’t answer anything – always run the pull command first.

🚀 Try it now

Open your browser, go to http://localhost:8080, and ask the chatbot “What’s a fun weekend activity for a rainy Saturday?”. In under two minutes you’ll see ChatRTX in action and know that your private AI is ready for the next task.

✦ Original step-by-step guide by AI World Co.'s AI editorial team. Written in plain language, reviewed for accuracy.

← Back to all stories