Transcription
If you're juggling AI tools, you're unsure what's actually worth paying for, or maybe you're worried you're missing out by sticking with one AI tool, well, this video is for you.
This is the ultimate guide to choosing the right AI. And by the end of this video, you're going to know exactly where to spend your time and money for your specific needs.
By the way, if you're new here, my name is Rick Mready, and for the past 11 and a half years, I've run a multi-sefigure online business helping other online businesses grow and scale their businesses. Today, I run a membership community called the AI Playbook, where I help online entrepreneurs leverage AI so that you can streamline your business, become more profitable in the process. I'll link to the community in the description below.
Now, for each of the five AIs I'm going to share with you here today, I'm going to share what I think each one is good for, what it's not great for, privacy concerns, if any, with that specific AI model or tool, limitations of the free versions of the tools, and then also whether I think it's worth upgrading from the free version.
And the first AI we're going to talk about is basically synonymous now with AI, and that is ChatGPT. ChatGPT is great for your day-to-day AI usage. It's versatile. It's really good at brainstorming, fleshing out ideas, analyzing documents, reasoning, and image generation. It's actually really good at image generation. Now, with the 4.0 image generation since OpenAI made the update to it, it's really good.
On the flip side, it's not great with large documents and inputs because of the smaller context window size of 128k tokens unless you're using the 4.1 model as that has a 1 million token context window. The other thing with ChatGPT is it tends to hallucinate quite a bit. So, as with any AI, you want to make sure that you're always checking the outputs for accuracy of what you're getting on a ChatGPT.
Now, in terms of privacy when it comes to ChatGPT, all of your chats, your conversations, anything you upload in the ChatGPT by default is being used to train their models. So, if you don't want that to be the case, you have to opt out of it.
Now, the other thing to remember though is if you connect to third-party apps, say like Google Docs, you are inheriting those third-party apps' privacy rules. A lot of people don't think about that.
Now, ChatGPT does have the temporary chat option, which means nothing that you say in the chat or a conversation when you have temporary chat turned on is being used to train the model. Likewise, it's not going to show up in your history of chat conversations in the left-hand side.
Now, when it comes to the free version of ChatGPT, there are some limitations you're going to want to be aware of. You're not going to have access to the more powerful models like 03 or 04 mini high or the full 4.1 version.
Now I say the full 4.1 version because when you're using the free version of ChatGPT you're using 4.0, and you get a certain allotment of using 4.0, and then if you go through that usage allotment of 4.0 then it will kick over to 4.1 mini, which is still a really strong model, but you don't have access to the full power of 4.1 like the million token context window.
Now you are also limited on the free plan to about 10 messages every 3 hours. You also have very limited usage of advanced voice mode. Image generation is capped daily. So if you are using a lot of images or you want to be generating a lot of images, the free version isn't going to be for you. Deep research, which is also really good, is also capped.
So is it worth upgrading from the free version of ChatGPT? Well, if you're more of a power user and you want access to the more powerful reasoning models like 03, the million context window of 4.1, you create a lot of images, you want to do a lot more research using deep research, if you want unencumbered use of advanced voice mode, which is great by the way, uh, also if you want to be creating GPTs, which you cannot do on the free version, then yes, I think ChatGPT paying for one, at least the $20 a month plan, is definitely worth it.
The next AI we're going to talk about is Claude. And literally 30 minutes before I started recording this video, Anthropic just dropped Claude 4, which we'll talk about briefly here in this section. But Claude is great for content creation, for nuance conversation, for ethical considerations, for coding, really good at coding, analysis, reasoning, and if you're on the paid version, which we'll talk more about, it has really good tie-ins with Google apps like Gmail, Calendar, uh, Google Drive.
Now, on the flip side, it does have a very limited context window in my opinion with 200,000 tokens, which can bring about a maddening stop to longer chats. It also has fewer multimodal capabilities than other models with just image.
Now, I was hoping that with the release of Claude 4, Anthropic would address these two things with the with a potential increase in a context window. That has not been the case, at least at the time of recording this video. And then also in terms of image generation, you still can't generate images within Claude.
Now, in terms of privacy concerns, this is where Claude really shines. It does not use your chats or conversations or anything you upload to train its model by default, and it's actually the only AI that does that. It is not turned on by default. Just like ChatGPT though, if you're connecting to third-party apps, like I mentioned with Google Drive or Google Calendar, Gmail, whatever, you are inheriting their privacy rules. So, keep that in mind.
Now, when it comes to the free version of Claude, you are pretty limited. It has quite a few limitations. The highest model you have access to is Claude 4 sonnet, which again is much better than the 3.7 sonnet model which is replacing, but you are going to run into daily usage caps on the free version. You don't have access to the new Opus 4 model, which is Anthropic's smartest and most powerful AI model yet. You don't have extended thinking mode. You won't have access to the web search tool. No Claude projects, which is a big thing I think. And you also will not have access to the Google Workspace, GitHub, uh, Dropbox connections that you do have on the paid versions.
So, is it worth upgrading from a free Claude plan to a paid Claude plan? Well, honestly, you're pretty limited on the free version. So, if you want to take full advantage of Claude's capabilities, especially with the new Claude 4 coding, content creation, complex problems that need deeper reasoning, if you want Claude projects, you're going to want to upgrade. And especially if you're a coder or doing any kind of coding, you're going to want to upgrade. But if you're somebody who generates a lot of images, for example, maybe you don't create a whole lot of content, you don't do much coding, maybe you're not a Google Apps user, I would say no on upgrading.
Google Gemini is the next AI we need to talk about, and I kind of have a love-hate relationship when it comes to Google Gemini. Uh, I'll talk more about that here in just a second. Now, it has a huge 1 million token context window, which isn't anything new. They've had that for a while. But what this means is you can work with large documents, really large documents and pieces of content, which is great. So this is going to be a huge differentiator for you. If you do work with large documents and need that context window size, Gemini is going to be really useful for you. It's also great with multimodal capabilities, coding, and I also have really good success with using it for coming up with ideas, brainstorming, fleshing out those ideas, and just really working through problems that come up during the day.
On the flip side, I do find some drawbacks when it comes to Gemini. Number one, it is not great for creating content for the most part. I also tend to find the responses that I get from it, I don't know, dry, I guess, is the best way to describe it. I'm like kind of lacking personality. Even when I give it my voice, it tends to kind of struggle with that. The other thing that drives me bonkers when it comes to Gemini, and this is so ironic, right, because it's Google's AI model, is that it doesn't always give you the most accurate information. Like I had one instance recently where I had to ask something four or five times just pushing it because I knew it was giving me incorrect information, and it took a lot of back and forth to actually get it to give me accurate information, and it was really frustrating.
Now in terms of privacy concerns, privacy when it comes to Gemini is governed by Google's general privacy policy and a specific Gemini apps privacy hub. Now, you do need to opt out of your chats, your voice recordings, uh your screen recordings, anything that you upload is going to be used to train the model. So, if you don't want that to be the case, you've got to opt out. And you can do that with the Gemini apps activity option. And similar to the other AI tools we talked about, when you're connecting Google apps to Gemini, it's all under the Google umbrella. So, it's all under Google's privacy rules.
Now, when it comes to the free version of Google Gemini, there are again some limitations. The highest model you're going to have access to is Gemini 2.5 Flash, which is a very good model. It's a very good day-to-day AI model. Uh, you are going to be capped though on document and image upload and also deep research. Deep research with Gemini is very good, but on the free version, it's going to be capped. You also are going to have lower usage caps in terms of just the number of chats and uh responses that you'll get from 2.5 Flash on the free version.
Now, do I think it's worth upgrading to a paid version, a paid plan of Google Gemini? Well, first and foremost, you do have the million token context window when it comes to Gemini. So, if you use huge documents and you're working with large amounts of content, then right off the bat, that automatically eliminates other models. So, what I would do is I would compare the paid version of Gemini to the paid version of ChatGPT because 4.1 now has the million token context window. So, I'd compare those two and see how you like the results between the two. If you are not worried about the context window, if you are a Google ecosystem user, meaning you use Google's apps, as Gemini becomes more and more integrated within Google apps and workflows, when that comes out and you're a Google ecosystem user, then I think yes, you're going to want to be on a Gemini paid plan. Now, if you are a Google Workspace user, and I think it's like the business plan and up, then you're already getting access to all the paid features of Gemini, which is great. If this does not describe you, if you're not a Google Apps person and you don't really care about the Google ecosystem, then I would not upgrade to the paid plan of Gemini.
The next AI that we need to talk about is Perplexity. I've got a love-hate relationship when it comes to Perplexity. It is great for fast search with citations. It's also really, really good for real-time, up-to-date information. So, if you need to be doing web searches, I talked to a lot of people who no longer use Google to do their searching, they go right to Perplexity. Now, you can also access several other AI models, the more advanced models, if you do a paid plan of Perplexity, which we'll talk about here coming up.
And on the flip side, I don't think that Perplexity does a great job, and I know that I'm in the minority here, but I just haven't seen it. I don't think it does a great job with depth of research. For quick searches with citations and up-to-date, cool. But depth of research, even with the deep research feature, I think it really lacks in depth of response. And I talked about that in another video I've done here on the channel when I talked about the different deep research models. So, if you're looking for really deep research on whatever it is that you're wanting to know more about, I don't know if Perplexity is the best tool for you.
Now, if you're looking for quick searches, you know, again with citations real-time, it's a great tool.
Now, when it comes to privacy for Perplexity, your usage and queries are logged, and that is on by default. So, you do have to go in and turn that off if you don't want any of your information and queries being used to train their model. You do have to go turn it off yourself. If you're a pro user, your data has a bit more privacy with it. Perplexity is also pretty clear about email and calendar syncing data is not used for training.
Now, what about the free version of Perplexity? Are there any limitations? Well, you are limited on the number of searches that you can do per day. I think it's three searches per day on the free plan. And then also, you are limited. It limits the number of sources that it will search for you on whatever query that you're putting in. So other than that, it is a pretty strong free plan. I will say you are limited on the options for the advanced AI models that you can access within Perplexity on a free plan. You have to upgrade for access to those. And then once you do upgrade, you it will search more sources, and also I love the social searching that it can do. So it can pull in Reddit and other social channels, channels as well.
Now, do I think it's worth the upgrade from the free version of Perplexity up to a paid plan? If you do a lot of research and you rely on up-to-date information that is accurate with citations that you can go and check the work, then absolutely yes, I do think Perplexity is a great AI tool just for the sheer fact of lifting the usage caps and also allowing you to search more sources. I think that alone is worth the upgrade to the pro plan for Perplexity.
And the fifth AI model to talk about here today is Grok. Now let's talk about what it's really good at. It's really good at real-time information, current events. So if you're searching for real-time, you want to know what's going on up-to-date, Grok is really good at that. It's also going to give you sort of an edgy kind of personality for an AI. So if that's something that you're looking for from an AI model, Grok is going to give that to you.
Now on the flip side, something to keep in mind when it comes to Grok, it is a pretty niche platform as its responses rely heavily on real-time information from X's data and content. Compared to the other AI models, Grok is not great for content creation, deep academic research beyond the content that it pulls off of X. It does have limited integrations, and you really want to be careful when it comes to factual accuracy. I mean, like with any of the AI models, based on the responses that it gives you because it's pulling so much data from the X platform, that can be a concern if that's important to you. It's also really not going to be your workhorse AI, say like a ChatGPT. With that said, I also know a ton of people that really like it and use it all the time.
Now, when it comes to privacy concerns for Grok, you do need to opt out. Just like many of the other models, you have to opt out of your chats being used to train the AI model. And to do that, you have to go into your X account in order to turn that off. Now, they do have the private chat option, which is similar to ChatGPT's temporary chat, and the same sort of thing goes there. So, if you haven't opted out of your chats being used to train the model, and you do want to have a cool private chat, then turn on the private chat option, and whatever you put in there is not going to be used to train the model.
Now, what about the free version of Grok and are there any limitations? Well, I will say that Grok has a very good, very strong free plan. You only get about 12 queries-ish per two hours, but considering how that compares to other free plans on other AI tools and models, that's pretty strong. You also are going to be limited in cap to sort of the deeper research and so forth that they have.
Now, is it worth the upgrade? Well, if you use X a lot and you value its real-time X data integration in your output and you like the distinct personality that comes with Grok, then yes, I do think it's worth upgrading. But I would do it as part of the X Premium package, with at the time of recording this is about 30, I think it's $30 a month for the X premium package. If you're not a big X user and you don't want the real-time X data integration in your AI outputs, then no, upgrading to a paid Grok version is not worth it for you.
All right, I hope this was helpful for you today. I think the big thing to remember is that there's no one-size-fits-all AI model or AI tool. It's whatever best fits for your specific use case on a day-to-day basis. If you said, "Rick, you have to choose one AI tool that you use day in and day out and you're going to pay for," for me personally, it's going to be ChatGPT. I really like it. I use it all the time. But I also subscribe to all these other ones because I love comparing the outputs of one model to another to another. And that's just the way that I think. I mean, that's my business, right? So, that's the way that I do it. But again, if you had to pick one model, if I had to pick one model for my day-to-day, I'm going to be choosing ChatGPT.
All right. Thank you so much for watching. As always, if you want to come check out my AI playbook community and get a free 7-day trial, I'll link to it in the description below.