OpenAI has just released the new GPT-4o model, which you can now try for free online.
That’s right, this is yet another new AI model to know about. However, there are a couple reasons why you should care about this one in particular. First, it’s the new ChatGPT model, a brand that remains the biggest generative AI bot around. Second, it generates audio along with images and text. And did we mention it’s free?
Here’s what to know about GPT-4o.
What GPT-4o Does
The name GPT-4o is a reference to the fact that this is a version of the fourth big ChatGPT model, with the “o” standing for “omni,” and presumably thrown in to hype up this model’s “all-encompassing” functionality.
Like all LLMs, GPT-4o is built to generate media based off of previous data that it has assimilated and reproduced.
🔎 Want to browse the web privately? 🌎 Or appear as if you're in another country?
Get a huge 86% off Surfshark with this special tech.co offer.
But this version has a few new tricks: Since it generates audio, it can handle more mediums than the standard text-and-image bot.
The paid-only version also boasts a voice assistant that can translate visual input into audio descriptions in real time. One demo video shows a vision-impaired pedestrian getting GPT-4o to describe ducks in a nearby pond and to help hail a taxi.
My favorite use of GPT-4o
AI already helps a lot with accessibility and this is another level pic.twitter.com/JXQvWGLsUw
— Kris Kashtanova (@icreatelife) May 14, 2024
What Can You Do For Free With ChatGPT?
With the new GPT-4o model, ChatGPT can generate new audio, images, and text in real time. However, free users are limited to a set number of prompts per hour, which curtails GPT-4o’s usefulness as an accessibility aid.
Still, if the model can deliver on what OpenAI promises, this is the most functionality that free users have had yet.
And it all comes with a rollout of even more features that were previously limited to ChatGPT Plus: Free users can now access the GPT Store, a marketplace for third-party apps built on top of ChatGPT that was previously reserved for paying users.
Also freshly free? A web-browsing tool, memory features, and an ability to upload files for ChatGPT to work with.
Not available: That GPT-4o voice assistant and everything it can do. It “jokes, chides, apologizes, pretends to blush, and knows how to deal with interruptions,” according to Axios coverage of the demo. It also handles voice translation. Plus, paid users get five times as many GPT-4o prompts as free users.
When Can I Use GPT-4o?
The new model is supposed to be available starting today. If it is, you’ll be able to log into the ChatGPT site and use it, just like you would for the most recent free model, GPT-3.5.
The paid version, with its voice assistant, will be rolling out in a few more weeks, according to OpenAI. Hey, that gives you plenty of time to brush up on your prompt knowledge before the next hype cycle in the still-ongoing AI model battle kicks off.
