Video Gamer is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
If you’re wondering how you can download, install and use Nvidia Chat with RTX, then we’re going to be going over everything you need to know about getting started with running Team Green’s LLM from your home PC.
How to download Nvidia Chat with RTX on PC
You’re going to need to head on over to the Nvidia Chat with RTX page and you’ll be able to download the LLM from there. It’s a hefty download, and it’s a good idea that you have a fairly competent SSD and RAM to pair with your graphics card. Expect the full thing to be even bigger when it releases…
We’ve recently had a look at how difficult it is to run an local LLM from your PC, which we’ve deduced is a much more difficult and irritating task than it seems.
How to install Nvidia Chat with RTX on PC
Thankfully, installing Nvidia Chat with RTX on your PC is a much easier affair than anything you might have previously tried. Once you’ve downloaded the massive zip file from Nvidia, you’re going to want to unzip it. That’s not going to be a fast process either.
After that, you’re going to install Chat with RTX as you would any other type of Nvidia driver installer. Click through the instructions and choose whether or not you’d like to do a fresh install. This is recommended if you’ve previously tried installing, and have not been successful.
How to use Nvidia Chat with RTX on PC
Using Chat with RTX is also pretty straightforward, and it comes with a few LLMs pre-downloaded. While using models such as Google Ultra or GPT-4 won’t be possible as they aren’t open source, others such as Mistral and Llama are readily available.
While Chat with RTX has the option for you to train your model on your own data, you’re also going to be able to use the sample model default data. While you’ll be able to ask it all the same questions as you would to ChatGPT or Google Gemini, don’t expect similar responses in terms of reasoning, clarity, and realism.
If you choose, you’re going to be able to provide Chat with RTX with a dataset. At the moment, simple filetypes such as .txt, .pdf, and .doc files are accepted, alongside YouTube video URLs.
Once you’ve ‘trained’ the chatbot on your data, you’re going to be able to ask it questions. As a journalist and writer, I’m already seeing opportunities in how this could streamline my workflow significantly: longform articles, endless datasets, and mind-numbing YouTube videos will all submit to my AI enhanced efficiency now.
Is it free to use Chat with RTX?
At the moment, Chat with RTX is still a tech demo, so it’s entirely free to use.
