ChatGPT Models and Pricing Plans in ChatGPT

All right, so let's talk about the different pricing plans and models that you get as part of this. So, in ChaiTPD there's different pre-trained transformer models that are the way that they affect the way that ChaiTPD can create responses. So, models affect the way that it creates that response.

How does it generate? What information does it know? And how does it craft those generative responses that it generates for you? So, the latest models, there's actually just a couple different ones, but this 4.0, which is an O as in Omni, not 4.0. They're really bad at naming things at OpenAI. It's really weird because they used to have ChaiTPD 3.5, and you'd think that 4.0 would be next. And they're like, oh, let's do 4.0 as in Omni.

Omni, like text, voice, vision, it's not just one thing, it's Omni. Like Omni directional things are like all over. So, it's like all these different ways to interact, not just chat.

For text, we could do voice, we can see. So, yeah, it's just a bad, yeah. It's not 4.0, it's 4.0 as in up.

So, these 4.0 models here are the latest ones. And then here's just to confuse you more, there was just GPT-4, plain 4. So, they had 3.5, which was like the older one, then they went to 4, and then 4.0. You guys are bad at naming things. And they have three different 4.0s. They have regular 4.0,4.0 with Canvas, and 4.0 Mini.

AI Classes & Bootcamps

• Live & Hands-on
• In NYC or Online
• Learn From Experts
• Free Retake
• Small Class Sizes
• 1-on-1 Bonus Training

Named a Top Bootcamp by Forbes, Fortune & Time Out

Learn More

So, basically, the regular 4, that was their older model. The 4.0 is their latest one. And when they have just a regular name, that's like their full-fledged model.

They often have a mini version. What's the deal with the mini versions? The mini ones are lighter weights. They might not have all the same features of the full-fledged model, but because they're lighter weight, they can run faster.

So, you can get faster responses. But also, these things are computationally complex. So, generally speaking, these newer models, they take more energy and they take more computing power.

And with all the GPT users out there, so all the chat and GPT users, if everybody was trying to do this all at the same time, they'd overload the computers. So, they have to limit how much access you get to these more advanced models. So, even when you pay, you do not get unlimited access to the latest, best models.

You get some access, but if you keep doing chats after chats after chats, eventually, they're gonna be like, okay, you're done for a little while. And you get a certain number of messages per month to these better models. Whereas the mini models, you get basically unlimited messages to those.

So, the mini models are there when you don't need to use the latest models for your stuff, when you might not need all of the features. And those are lighter weight, you're gonna get faster responses, and you're gonna get more messages to those. So, we don't like kind of wasting our messages to the better models when we can use the minis instead.

And the 4, that's even older. So, again, that's kind of like, we kind of get more messages to that. So, if, let's say, you lose your messages to 4.0 and you have to fall back to 4, they keep those older models around just because generally they're easier to compute and they can kick you back to those older models once you've used up your messages to the newer models.

If you're in a free account, you don't choose your models. In a free account, they just say ChatGPT, or if you want Plus, Plus is what they call the paid membership, you can get a ChatGPT Plus membership where you can get the latest, smartest models. The regular ChatGPT, they say it's great for everyday tasks.

What that is doing behind the scenes is they're gonna default you to 4.0 because they want you to see their latest and greatest thing. And then if you use up all your messages to that, then you kick back to 4.0 mini, basically. But you don't get to choose when you create your chat in a free account.

So, if I say, you know, tell me anything you can do, this is a basic thing. This does not require the latest 4.0 model to be able to answer that because it's a very basic question. And it'll go through, it'll do the answer, but I didn't get to choose to use the mini.

So, in a way, I kind of wasted this question on 4.0, which didn't, I didn't need this level there. Afterwards, can I switch to another model? Yes. In the free account, afterwards, you can switch.

So, by default, it basically does it automatic. But I know if I need more features, I can use 4.0. But if I don't need them, I can switch to 4.0 mini. And that, you're gonna basically get unlimited ones for the 4.0 mini.

And you won't waste your messages to 4.0. But you don't get to choose that in a free account from the start. You'll kind of burn one of your 4.0 messages at least. And then you can switch back to mini.

And you can use the mini to do it. And so, from now on, this chat will use the mini model, which will be faster. Like that actually generated faster.

And I didn't need the advanced features of 4.0. The mini will be faster. And I get unlimited messages to that. Now, in a paid account, when I create a new message here, in a paid account, I get access to all the models.

So, I can choose 4.0. I can do the 01 with Canvas. I'll talk about all of these models here. I can go back to the 4.0 mini.

Notice I even kind of pooh-poohed that. They're like, oh, we're going to give you the latest, greatest stuff out here. But yeah, you can go back to the mini if you want.

Or you can go back to the really old 4.0. Not that I really do that much. So, what's the difference between 4.0 and 4.0 with Canvas? Let's talk about that. So, the Canvas part, if you think about, if you've ever done Google Docs, you know how when you're in Google Docs, you and other people can go in.

You can all be kind of working on things at the same time. So, think of this as like you and ChatGBT working on a Canvas together. So, you can kind of have a document and be going back and asking questions.

And it's more of an interactive Canvas to work on, that you can actually work on a document here. Now, this only works in paid accounts, but I can still show you how this works. So, let's say I wanted to write an article on something.

If I just use 4.0, it's going to produce an article and I can't really interact with that article per se. So, let's say I say, write an article on, I'm going to say, write a 500 word article on evolution. Okay.

So, it'll write that article here. Right now, it's using 4.0 to do this. And once I'm done, it's kind of, it's just got to be done.

I could write follow-up questions to have it enhance this, do things. But this thing is kind of done. If I ask it to do something, it'll create another response down below and another response down below.

But if I took that same question, it's not going to do it. And if instead, when I create my chat, if I do it with the Canvas, notice I say beta, this is one of the newest ones. It's required as a paid account.

So, now, this is the Canvas that we have here. So, now we've got a chat kind of on the side here. So, now, let's say, if I don't like this.

So, well, first of all, I can say, like, maybe over time, maybe this. I'm like, okay, let's bold that word. Because maybe I like that, you know.

So, first of all, then I can even interact with it at all and do things. Another important, I can actually go in and edit that. It's editable.

So, maybe I don't like their wording. So, now, I could follow up. I could say, I could interact over here.

But I could say, please add headings and use bulleted lists more. So, I could have it just edit it. Right? And notice how it does that.

So, it could be that I give a follow-up thing and then it rewrites it. Right? But, oh, see, notice, remember how I folded that? So, it kept that. So, it still kept that and did what I said.

So, this is a more interactive canvas to be working on. Because it's like the two of us, my digital friend, Chachapiti, and myself, we're collaborating on something together. And I'm having Git work with something.

Now, maybe I wanted to maybe here there's a little message thing. I can give it a message and say, please add a second example. And I missed.

Please add a second example. So, now, it's editing that. And so, now, they just added a second example.

Where I wanted it. So, the canvas part adds this interactive, almost like Google Doc collaborative thing. Instead of me just having it generate response after response, I can go specifically in areas and say rewrite this area.

Like I could say for this, rewrite this with more about geography. I don't know if that makes sense. So, now, go back and rewrite about that.

Right? So, that's the canvas part of this. Which this model is a paid model. So, in the regular so, going back to the different models here.

In the free version, you get 4.0 and 4.0 mini. Those are the two you get. You can't choose to start you will be defaulted into 4.0. You can switch to 4.0 mini.

But what will happen is once you run out of 4.0 messages, you can only use mini. And that means you get even fewer features. Because 4.0 has more features than mini.

So, there's things like you can't upload files. Can't browse the internet. Doesn't have vision capabilities.

Can't do data analysis like you upload a PDF and have it do data analysis. There's many features that 4.0 can do that the mini version can't. So, they kind of want you for free to waste your messages to 4.0 because they want you to realize that mini sucks and they want you to pay to upgrade so that you'll get more messages that you can do to 4.0. Okay.

So, now, let's finish talking about models that we're going to get into pricing and all that sort of stuff. Okay. So, there's also 01 models.

And I know they suck at naming things, right? Because 01, is that the first? No, that's the latest. This is only accessible in the paid version. So, they also have these 01, right? So, they have the 4.0 and the 01.

Could you name things any worse? And they have a preview release because they're previewing. This is like beta testing. The mini is faster.

Doesn't do as much, but it's faster. The preview, this is the full-fledged model, can do the most stuff. So, anytime they have minis, you get kind of more messages to mini.

It's faster. But it doesn't do as much as the full-fledged model. So, these are very different in the way that they work.

So, the way that the regular ChatGPT models work is you send a message and it just starts writing a response. 01 models, they're slower because they have to think before the answer. They have to look at all this stuff and they'll actually go through this internal chain of thought process.

So, if somebody asks you a question, you might say, well, first I need to do some research. I need to evaluate potential answers. I then need to choose the correct answer.

In a regular chat, if you just start writing a chat and it just starts writing a response, it didn't evaluate all the potential responses that it could do. So, if you need more advanced kind of high reasoning, things like, for example, physics, chemistry, biology, those things, which need more thought process, these are built for that higher level thinking. Because it's going to take longer, but it's going to think through its response more.

So, it's trying to be more like a reasoning person. Now, if you're asking it a basic question, you don't need this advanced reasoning for this in every single case. So, the 01 preview, this is their full-fledged one, it's for hard problems.

Like, things that are not easy to answer, where you need to really think about something. The mini, it's faster and cheaper. It's also, but it's still one of these reasoning models.

It's really good at coding. So, any sort of coding language. Because you don't want to just rush into some code, you want to make sure it's the best code for that job.

So, you might need to look at some alternatives. For math, math, to get that right, you can't just start writing a response. Like, you need to think about, is this the correct thing? Same thing for science.

Now, these are the latest models. They only give you, for the regular 01, they only give you 50 messages per week. Even in the paid ones.

And the mini, you get 50 per day. So, in the 4.0, you're going to get more than that. This is the most limited of all of these.

But you think about 50 per week for this. Or 50 per day for this one. So, but, you know, if you're doing coding, or if you're doing math or science, you know, 50 per day, depends on how much you're doing.

Even for the paid ones, you still might run out of that, depending on how much you're using. Because remember, all your follow-ups, your try-agains, right? Because it might not get it right the first time. So, each one of those is another message.

So, you could chew through 50 per day pretty quickly for these more advanced things. But not everything needs this. So, this is for the harder, more complex things.

And that's why they don't default you to that. They default you to 4.0, because that's kind of general intelligence. But when it comes to coding, math, and science, those kind of things, or anything that is hard to answer things.

Things you need to reason and think about, and take some time to think about. Those are things that the 4.0.1 models are going to be better. Those are advanced reasons.

For users, because you get limited access to the 4.0 stuff, you're going to get limited access to these more 4.0 features. So, we're going to go through these features, but things like data analysis, to be able to upload files, these are all features of 4.0. So, you can do some of this. But once you run out of those 4.0 messages, and you get kicked back to the mini, the mini version of it can't do this stuff.

So, basically, you just keep using mini for free, or you can pick. Now, you get a limited number of messages in a three-hour window for the free level. So, in the free level, if you're messaging it a lot in three hours, they might say, hey, come back in like two hours.

And they'll tell you, they'll come back in two hours where you can do more 4.0 messages. Or in the meantime, kick back to mini for free. So, they reset every three hours.

So, today in class, sometimes people will hit that limit, depending on how many messages you say to the 4.0. You might hit that because it's a six-hour class today. So, people will hit it, and then maybe they'll say, oh, come back in an hour and a half, come back in two hours, and you'll get more 4.0 messages. Or in the meantime, you can use the mini, or you can pay to upgrade, and then you'll get more 4.0 messages.

So, they don't say the exact number of messages, just that you get limited. And they say the paid one gets five times the amount. But they're not specific of how many.

Partially because I think it depends on how overworked their servers are. So, if they're being really worked a lot, they're like, hey, we're going to give free users less. And even paid users, you know, they might have to vary a little bit, depending on how overloaded they are.

There's also an idea of context. Context is how much can it understand? How much can it kind of keep in memory? For example, in 4.0 models, you can upload files. So, you could say, summarize this document.

Now, this could be a book, it could be an article, it could be some business document, it could be a legal document. So, when you have a bigger context, it means more memory, right? So, you can have bigger files. So, let's say with a larger context, maybe it could be a novel.

But maybe with a smaller context, maybe it could be a certain number of articles, right? Versus like a full novel. So, imagine if you wanted to upload a novel and say, tell me, like, summarize this novel. But if you don't have a big enough context that it can't read all of that novel, because it doesn't have enough memory, doesn't have enough context, it can't summarize it.

You have to break it up into chunks. But then it might not be able to remember what happened in the first half when it gets to the second half. So, it might miss the point of things.

This is where having bigger context helps you to upload bigger stuff. But still, it's not like small context is small. You can still upload quite a bit of stuff to it.

But the bigger the context, the bigger the files, the more stuff you can give it and remember all that stuff. Also, as you go through a chat and you fill up that chat, chat message after chat message, all of that is adding to context as well. So, the longer or bigger your context is, the more it can remember in a particular chat and the bigger files you can upload.

So, for free, they choose the model for you. You get access, whatever you want to the 4.0 mini and limited to the 4.0. So, therefore, because you're limited with 4.0, you get limited access to some of the more advanced features and they're going to give you a smaller context window. When you pay, that's ChatGPD+.

That's what they're trying to upgrade you to, to upgrade. So, that, you get access to all of the models. So, you get the older 4, the newer 4.0, you get the 4.0 mini, 0.1, you get all of those different models and you get five times more messages for, compared to the free plan.

They don't say how many messages, they just say you get five times the amount of messages. They just aren't specific. But because you get a lot more messages to 4.0, you get much more access to all the advanced features that we're going to be going through, things like data analysis, more file uploads.

They can browse the web. You can generate more images because for the free one, you have limited access to 4.0. 4.0, you can't do as much stuff with. So, there's many other features.

We'll be going through these features and talk to you about that. And you get a larger context window. So, because you get that larger context window, you can upload bigger files.

This is what they've latest set as a main. This could always, this subject's changed. Plus users can send 80 messages every three hours, plus users are the paid users.

So, 80 messages every three hours on 4.0. And the older 4, just the regular 4, they could do 40 messages every three hours. So, 4.0 still has some capabilities. So, if you run out of 4.0, you run out of 4.0 mini, basically they have those older models that you can always go back to if you start running out of the more advanced ones.

So, normally I choose one of the later models. And then if I run out of message to those, I can always fall back to the older models. But they do say that during peak usage, they might limit those.

So, these are kind of general ideas of about how many messages. So, if you're a plus user and you can send 80 messages, you can imagine as a free user, you don't get a whole lot of messages because it's like five times what the free users get. So, free users don't get a whole lot of messages to the newer models.

Peak hours are whatever hours people are using. Peak is based on how many users are hit. So, remember this is a worldwide usage, not just US.

So, anytime there's peak demand, if all of a sudden like maybe it's a holiday and everybody's hopping on ChatGPT or maybe it's a workday or, you know. So, whenever that peak happens to be. Now, they also have business plans.

The plus plan is really for individuals. The business plans require two or more users. So, this is for teams.

This is for companies that want to have many people on the same team, essentially. So, if you have two or more users, you get everything you have at plus. They give you unlimited access to everything in 4.0 mini and you get even more message limits to all the other ones.

Also, we're going to be talking about your data being excluded because that's something that we want to talk about is do they train on your data? Because by default, they do. Actually, what you put in there, they start learning on that and building it into their algorithm so that they can use those responses. So, don't worry, we're going to disable that.

By default, business plans opt you out of that by default. So, because they know that businesses don't want ChatGPT training on their data by default. So, in the free and plus plan, you have to opt out.

I'm like, hey, I was paying, you shouldn't train on my data by default, but they still do. They opt you in automatically. We can opt out, even in the free version, we can opt out of that, which we'll do in just a moment.

But by default, any business plan, they opt you out by default. You get that larger context that you get in plus, and you have to go up to enterprise to get an even larger context and to get unlimited. So, basically, you have to get all the way up to enterprise if you want to have unlimited messages.

So, and then they get more advanced stuff, enterprise verification stuff. So, if you're a company and you're having employees sign up, you can get in the team plan, which is basically the plus with some more messages. But if you're an individual, you would generally be signing up for the plus because you can do that one at a time.

You can just be a single user on plus. And this link here, anything that has this colored text, those are links, and you can go out and check. So, plus is $20 a month.

And the team is $25 per person a month if you bill it annually. Like, say, I'm going to do it for a year. If you do it month to month, they charge you more per month.

So, this one requires like a year commitment to do it. And that's per person. So, team, you'd have to be at least $50 a month for the two of you.

And for enterprise, you have to contact them to see how much they're charging for those. So, basically, for individuals, this is where most people start, for the big plan. So, $20 a month.

Which is the same thing as Microsoft Copilot. They charge $30 a month if you do a Copilot. These are just screenshots just showing.

So, in case you want to check out these later. So, if you don't have a paid account, these are showing you the models of the paid account. And you can do it when you create a chat.

You can start choosing your models or you can change it later. Whereas in the free account, you can't change your model to start. You can only change it later on.