Step-by-step
🎮 ChatRTX
NVIDIA
📘 Step-by-step 📘 NVIDIA · ChatRTX↗ ChatRTX · Official sitebeginner 🏠 Everyday life

Getting started with NVIDIA ChatRTX: a beginner's guide

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

6 min read·June 25, 2026

This guide was written by AI. It passed automated fact and quality checks; no human editor reviewed it.

By the end of this guide, you’ll have your very own private AI chat assistant running directly on your Windows computer, ready to answer questions, draft emails, or brainstorm ideas. It’s perfect for anyone curious about using AI without sending their data to the cloud, even if you’ve never installed complex software before.

💡 Tip: tap a step’s number when you finish it — a green tick appears and your browser remembers how far you got.

✅ Before you start
  • A Windows PC with an NVIDIA RTX graphics card: ChatRTX needs specific hardware (an NVIDIA RTX graphics processing unit, or GPU, which is the computer component that handles complex graphics and AI calculations). Look for an RTX 30 Series or newer, typically with 16GB of VRAM (the memory on your graphics card itself) for the best experience.
  • An internet connection: You'll need this to download the software and the AI brain, called a "model."
  • A free NVIDIA account: You'll need one to access the official ChatRTX download page.
  • About 30-60 minutes: This includes download and installation time.
1
Step 1 of 6

Create an NVIDIA account

Your first step is to visit the official NVIDIA website and create a free account. Open your web browser (like Chrome, Edge, or Firefox) and go to nvidia.com. Look for a button or link labeled "Account" or "Sign Up," usually located in the top-right corner of the page. Click on it to start the registration process.

You'll see a form asking for your personal details. Fill in your name, email address, and choose a strong password. This account acts like your personal pass to access NVIDIA's tools and downloads. After submitting, you'll usually get an email to confirm your address; click the link in that email to activate your account.

💬 ExampleName: Jane Doe
Email: [email protected]
Password: (choose a strong and unique one)
✅ You'll know it worked when

you receive an email confirmation and can successfully log into your new NVIDIA account.

Next step ↓
Create an NVIDIA account
2
Step 2 of 6

Download the ChatRTX server package

Now that you have an NVIDIA account, you can download the ChatRTX software. Log into your NVIDIA account on nvidia.com. Once logged in, navigate to the "AI Tools" or "Developer" section of the website. You might need to use the website's search bar to find "ChatRTX" if it's not immediately visible.

When you find the ChatRTX product page, look for a prominent "Download" button. Click this button, and your browser will begin downloading a large compressed folder (often called a "zip file," which is a collection of files bundled together to save space). This file usually lands in your computer's "Downloads" folder by default.

💬 ExampleIf you see a button that says "Download ChatRTX Server" or similar, click it. The file will be named something like `chatrtx_server_xxx.zip`.
✅ You'll know it worked when

you see a chatrtx_server_xxx.zip file (where 'xxx' are version numbers) appear in your computer's Downloads folder.

Next step ↓
Download the ChatRTX server package
3
Step 3 of 6

Install the ChatRTX server software

Before you can run the AI, you need to install its core software. First, locate the chatrtx_server_xxx.zip file you downloaded in your Downloads folder. Double-click it to open the compressed folder, then drag its entire contents to a simple location on your C: drive, like C:\ChatRTX.

Inside this new C:\ChatRTX folder, you'll find a file named install.bat. Right-click on install.bat and select "Run as administrator." This launches a command prompt (a text-based window where your computer shows what it's doing) and automatically checks your system. It also sets up necessary "libraries" (extra pieces of software the AI needs to function).

💬 ExampleC:\ChatRTX> install.bat
Checking GPU driver... OK
Setting up runtime... OK
🔄 If it looks different

or you see an error, ensure you extracted all files and are running install.bat as an administrator.

✅ You'll know it worked when

the command prompt window shows messages confirming that your GPU (graphics card) driver and the AI's "runtime" (the background program that makes the AI model run) are ready, and then closes itself.

Next step ↓
Install the ChatRTX server software
4
Step 4 of 6

Download the AI model

Now it's time to get the "brain" of your chatbot. The brain is called an "AI model," specifically a Large Language Model (LLM), which is like the powerful engine behind tools such as ChatGPT. To get this model, you need to open your computer's Command Prompt. Search for "cmd" in your Windows search bar and click "Command Prompt" to open it.

In the Command Prompt window, type cd C:\ChatRTX and press Enter to navigate to the ChatRTX folder. Then, type the command chatrtx pull --model default and press Enter again. The tool will begin downloading the AI model files, which are about 3-4 gigabytes (GB – a unit of digital information, like a large movie file) onto your PC. This process only happens once.

💬 ExampleC:\> cd C:\ChatRTX
C:\ChatRTX> chatrtx pull --model default
Downloading model... [===============================>] 100% (3.2 GB)
🔄 If it looks different

or you see a "command not found" error, ensure you are in the C:\ChatRTX directory in the Command Prompt.

✅ You'll know it worked when

the Command Prompt shows a progress bar reaching 100% and then confirms the model download is complete.

Next step ↓
ChatRTXCommand Prompt
✏️ Sketch — your screen may look a little different
5
Step 5 of 6

Start the chatbot locally

With the model downloaded, you can now launch your private AI assistant. In the same Command Prompt window (or open a new one and cd C:\ChatRTX again), type the command chatrtx run --port 8080 and press Enter. This command starts the ChatRTX "server" (a program that quietly runs in the background, waiting for requests) on your computer.

The server will begin "listening" on "port 8080" (think of a port as a specific "door" on your computer that programs use to communicate). You’ll see a line appear that says something like "ChatRTX ready at http://localhost:8080." "Localhost" simply means "this computer," indicating the AI is running directly on your machine.

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

Keep this window open; closing it will stop the AI.

🔄 If it looks different

and you get an error about "port in use," try chatrtx run --port 8081 instead.

✅ You'll know it worked when

the Command Prompt displays the ChatRTX is now running at http://localhost:8080 message.

Next step ↓
ChatRTXEnter
✏️ Sketch — your screen may look a little different
6
Step 6 of 6

Talk to your new AI assistant

Your private AI is now ready to chat! Open any web browser on your computer (like Chrome, Microsoft Edge, or Firefox). In the address bar at the top of the browser window, type http://localhost:8080 and press Enter.

You'll see a simple chat interface appear, much like popular online chatbots. This is your private ChatRTX assistant! Click into the text box at the bottom of the screen and type a question, like "What's a quick recipe for banana bread?" then hit Enter. The AI will process your request using its local model and reply directly in the chat window.

💬 ExampleYou: What's a quick recipe for banana bread?
ChatRTX: Here's a simple 3-ingredient recipe using overripe bananas...
✅ You'll know it worked when

ChatRTX responds to your question with a helpful answer in the browser window.

ChatRTXEnter
✏️ Sketch — your screen may look a little different
⚠️ Common mistakes
  • Forgetting to run the installer as an administrator: The install.bat file needs special permissions to set up drivers and software components correctly. If you get errors during Step 3, try right-clicking install.bat and choosing "Run as administrator" from the menu.
  • Using the wrong port or a port already in use: If you see an error about "address already in use" during Step 5, it means another program is already using port 8080. Simply change the command to chatrtx run --port 8081 (or 8082, etc.) and then open http://localhost:8081 (or your chosen port) in your web browser instead.
  • Skipping the AI model download: Without the Large Language Model (LLM) downloaded in Step 4, the ChatRTX server will start, but it won't have a "brain" and won't be able to answer any questions. Always ensure you run the chatrtx pull --model default command first.
🚀 Try it now

Open your web browser, navigate to http://localhost:8080, and ask your new ChatRTX assistant, "What's a fun weekend activity for a rainy Saturday?" In under two minutes, you'll see your private AI in action and know that your secure, local assistant is ready for your next task.

🎯 Next guideDraft Marketing Copy Instantly with NVIDIA ChatRTXLearn how to use NVIDIA ChatRTX to quickly generate product descriptions, ad headlines, and social media posts right from your computer.

❓ Quick questions

How long does this take?

About 6 minutes — the guide has 6 steps, and you can tick each one off as you go.

Which tool do I need?

This guide uses NVIDIA ChatRTX — but the approach works very similarly in other AI assistants.

Do I need to prepare anything?
  • A Windows PC with an NVIDIA RTX graphics card: ChatRTX needs specific hardware (an NVIDIA RTX graphics processing unit, or GPU, which is the computer component that handles complex graphics and AI calculations). Look for an RTX 30 Series or newer, typically with 16GB of VRAM (the memory on your graphics card itself) for the best experience.
  • An internet connection: You'll need this to download the software and the AI brain, called a "model."
  • A free NVIDIA account: You'll need one to access the official ChatRTX download page.
  • About 30-60 minutes: This includes download and installation time.
What mistakes should I avoid?
  • Forgetting to run the installer as an administrator: The install.bat file needs special permissions to set up drivers and software components correctly. If you get errors during Step 3, try right-clicking install.bat and choosing "Run as administrator" from the menu.
  • Using the wrong port or a port already in use: If you see an error about "address already in use" during Step 5, it means another program is already using port 8080. Simply change the command to chatrtx run --port 8081 (or 8082, etc.) and then open http://localhost:8081 (or your chosen port) in your web browser instead.
  • Skipping the AI model download: Without the Large Language Model (LLM) downloaded in Step 4, the ChatRTX server will start, but it won't have a "brain" and won't be able to answer any questions. Always ensure you run the chatrtx pull --model default command first.

Keep reading

📘 Chat with Your Own Documents Offline Using NVIDIA ChatRTX Learn how to privately ask questions about your personal files directly on your Windows computer, no internet needed, with NVIDIA's local AI chat tool. 📘 Set up a home-assistant chatbot with NVIDIA ChatRTX on your PC Transform your powerful NVIDIA PC into a private AI assistant for household questions, managing documents, and personal journaling, all without an internet connection. 📘 Draft Marketing Copy Instantly with NVIDIA ChatRTX Learn how to use NVIDIA ChatRTX to quickly generate product descriptions, ad headlines, and social media posts right from your computer. 📘 Practice Job Interview Questions with ChatRTX Like a Private Coach Run a realistic mock interview on your own PC, then get blunt feedback on your answers before the real thing. 📘 Turn Messy Meeting Notes into Clear Action Items with ChatRTX Drop your raw Zoom notes into NVIDIA's free, offline AI assistant and pull out a clean list of decisions, owners, and deadlines in minutes. 📘 Build a recipe generator that only uses what you already own with NVIDIA ChatRTX Turn your fridge into a meal-planning helper by teaching an AI to cook only with ingredients you already have, using NVIDIA's free tool.
Was this helpful?

✦ Generated by AI in AI World HQ's automated newsroom, written in plain language. Checked by automated fact and quality gates — no human editor reviewed this guide. Spot a mistake? Use the buttons above.

📘 Prefer it in one file? First Steps with AI: The Beginner Pack (12 guides, free printable PDF)

← Back to all stories