Transcription
If you're still juggling AI tools in 2026, maybe you're unsure what's actually worth paying for. Well, this video is for you.
Today, I'm going to walk you through four AI tools: Chatyppt, Claude, Gemini, and Perplexity. And for each of these tools, I'm going to cover what each one is great at, what each one is not so great at, any privacy concerns, free plan limitations, and then also whether upgrading to a paid tier with that specific tool is worth it.
Okay, so let's start with ChatBT. Chhatbt is good for day-to-day use, things like brainstorming, analyzing documents, reasoning, and now it's much better at image generation. And when you go into ChatBT, it defaults to the 5.2 model. And then you've got different modes here. The auto mode, which decides how long to think, has instant and thinking if you're on the pro version of Chachebt, and you've also got access to the legacy models, which are the older models here in ChachiBT. So you have 5.1 all the way down through 04 mini if you prefer to use some of the older models.
Now the context window with GPT 5.2 is now 400,000 tokens, which means you can work with larger documents and a lot more information at a time. And my experience with 5.2 thinking is that it tends to do a better job than past models when working with different large bits of information and a lot of different, uh, longer chats and so forth. But still, this context window size is still dwarfed by Gemini with its 1 million token context window.
Now, here at the chat window, you'll see the plus button, and this is where they have kind of a lot under the hood, if you will. This is where you can add photos and files and connect to your Google Drive, uh, create images, which I'll show you more in just a second. This is where deep research lives. This is where agent mode lives. Then you've got other options here like study and learn, web search, canvas. They now call their basically connectors or their MCPs. They now call them apps. So you can scroll down here and see what their featured apps are. You can click over to productivity and so forth and just scroll down there. There are quite a few here, but not as many say as in Claude.
The other thing too is if you don't see a connector here or quote MCP, you can go up to the top right corner and click on the little gear. And if you go down to advanced settings and click on developer mode, you'll notice that you can see create app. So you can create your own or you can bring in an MCP server, uh, URL or ooth that you want to be using that you don't see in the list here. As always, if you're using some random MCP server that somebody built, just be cautious of that, uh, for privacy reasons for yourself.
Now, another thing you'll notice in the drop down here is the create image option. You can also see it up here in the left column. I'll just click on that right now. You can describe a new image, try a style on for an image, and scroll through. Uh, it's way, way better than their previous image generation, which was not good at all.
In terms of limitations when it comes to ChachiBT, for me, when using the thinking mode here, so I've got this prompt in here, uh, I'll show you. It tends to take a very long time. It's slow. I'm talking like 45 plus seconds for responses that feel like they should take half of that. And I've experienced that since 5.2 came out. And this is only the standard thinking mode, by the way. You can click the drop down here and click on extended. Now, I'm just talking the standard thinking mode. It's slow. And also, ever since ChatBT 5 came out, even with the the newer 5.2 model, I find it to be much more rigid in how specifically, how so specifically you need to prompt it. Now, the outputs from ChachiBT are just not as good as they used to be with previous models.
Now, when it comes to privacy, I am skeptical when it comes to privacy for all of the AI models. But by default with ChatBT, your chats and uploads are going to be used to train the model unless you opt out, which is you just go down here to your name in the bottom left-hand corner, click on settings, and then click on data controls, and you'll see here improve the model for everyone, and you just flip that to off. It will default to on. If you connect third-party apps like Google Docs, for example, their privacy rules apply. The apps that you are connecting to, their privacy rules apply.
Now, each of the AI tools I'm sharing with you today do have free plans. So, when it comes to ChachiBT, no access to advanced reasoning like 5.2 thinking like I just showed you. You do have message limits, which is roughly about 10 every 3 hours. You've got limited voice mode, capped image generation, limited deep research, and you don't have access to custom GPTs.
So, with all this said, is it worth upgrading to a paid tier? I would say yes, it is worth upgrading if you want to be creating custom GPTs. Otherwise, they're just better AIs worth your money in my opinion.
In addition to these AI models, people also keep asking me which AI tool will actually matter here in 2026. Well, the one that I use daily is Recall. The problem is that we consume endless content, right? Podcasts, articles, YouTube videos, etc., but retain almost nothing. Well, the solution, Recall, is your personal AI encyclopedia that remembers everything for you. Save anything with one click. YouTube videos, podcasts, PDFs, articles, even recipes. Recall instantly summarizes and organizes it all. No tagging, no folders, no mental overload. And my favorite feature is the ability to chat with your entire knowledge base like it is your second brain. Like I can type in, pull insights from the business podcast that I saved, combine them with all my personal goals, and tell me what to focus on next. Done. And with Graphu 2.0, Recall automatically maps how all your saved content connects, revealing patterns that you'd miss on your own. Over 500,000 professionals are already using Recall to learn faster and think clearer. So, I want to invite you to try Recall for free at getrecall.ai or use code rick25 this January for 25% off the unlimited version.
Okay, now let's talk about Claude. Claude shines in content creation. Hands down for me, it is the best AI model for creating content. It's really good, but also with nuance conversation, ethical reasoning, analysis, coding. Claude code is amazing, and Anthropic is really building Claude to be the business go-to AI tool and models. And similar to what we saw in Chatbt, they have a lot of options here. So again, yeah, you can add files and photos, you can take a screenshot, which is pretty cool. You can take what you're doing here in the chat and add it to a specific project, a Claude project that you have. Uh, again, you can also add from Google Drive. You can also click on the search field. You can also add things directly from Notion because I have it connected to Notion, which you can create your own writing style right here. Also, the connectors or again, MCP connections to other tools is pretty robust here in Claude, which I'll show you more here in just a second.
Then also with Claude, similar to Chatbt, has a desktop app, but Claude's desktop app for me is where Claude really shines. For example, the ability for non-technical people like myself, honestly, to easily connect to files on your computer and combine that with say, for example, Claude projects. It's a game changer. So you'll see here under connectors, you've got web version. So you can scroll down here and it'll connect to say like N8N, GoDaddy, uh, Asana, etc. Then you've got specific desktop app extensions that you can connect to like control Chrome from within here. You've got the file system extensions. So again, you can be reading and creating new files and searching your hard drive, uh, through Claude if that's something that you want to do. I have this set up and it's awesome.
Also, as you can see here in the left-hand column, you've got artifacts. So you have the ability to quickly create apps without coding in any way, uh, in Claude's artifact. So, you can use one of their ones that they have already. Or you can click on new artifact and start creating your own from scratch if you want to build something out really quickly and create some sort of an MVP, if you will.
Now, another really cool feature of Claude that has to be mentioned as we're talking about it here is Claude skills. I've done a whole video about it here in the channel, but if you go under your, uh, your icon, go to settings, and then you click on capabilities, scroll down just a little bit, and you're going to see skills down here. So, if you already have skills, then you're going to see them right here. Otherwise, you're going to see example skills. Now, these are skills that, uh, Anthropic has already created for you. You can just turn them on, turn them off as if you want to use them. But I'd highly recommend that you turn on skill creator because you can just ask Claude in the chat, you know, I want to create a skill to accomplish, you know, X, Y, and Z. Please help me create it. And it will walk you through creating that skill.
Now, in terms of its biggest drawbacks when it comes to Claude, relatively small context window, about 200,000 tokens within Claude itself. If you're using the API and you're a beta user, you can have up to a million token context window. And it's also limited at this point with multimodal capabilities. It still can't generate images, for example.
Now, in terms of privacy, Anthropic, out of all the models, Anthropic does tend to focus more on responsible AI. It is now defaulted to on to help improve Claude, if you will, through your chats and conversations. To turn that off, you have to go into privacy, and at the bottom here, you'll say help improve Claude. You want to turn that off if you don't want your chats and conversations to be used to be training the model.
Now, when it comes to the limitations of the free plan with, uh, with Claude, you are going to have daily usage limits. You're not going to have access to Opus 4.5 or extended thinking mode like we talked about. You're not going to get access to Claude projects. You're going to have very limited access to connectors. If you're a light AI user, the free plan might be great for you. However, I do think it's worth upgrading, especially if you are a content creator, if you're a coder, if you are somebody who's solving complex problems, then yes, I absolutely think it's worth upgrading.
Okay, now let's talk about Gemini. Gemini's 3 Pro model is really, really good. It's actually rumored that OpenAI moved up the release of GPT 5.2 by a couple to a few months because when Google released Gemini 3 Pro in Q4 2025, it surprised everybody about how good it was. I used to use Gemini really rarely. I barely used it, but now I find myself using Gemini testing outputs and so forth almost daily.
Gemini's got a massive 1 million token context window, making it really, really good for working with large documents, long conversations. It's very strong, very strong at multimodal tasks. So, image generation, which I'll show you here in just a second, video, audio, etc. It's really good with coding, brainstorming, problem solving. Not as strong as Claude, for example, with content creation that I found, but overall really solid model.
You'll see here in the chat box, you've got a little drop down here where you can access, uh, the quote unquote fast version of Gemini 3 thinking and then their pro version. I do get access to, which I'll talk about more here in just a second. I get access to Gemini Pro through my Google Workspace account. And with that, you get access to a lot of the paid features and AI features and tools that Google does. Gemini, for example, you get the upgraded version of Notebook LM, etc., etc.
Its deep research is also my go-to. You can click on the sources. So you can choose one or more sources, search, Gmail, Drive, Chat, if you're connected to those tools here. You can upload files, etc.
Google's also done a really good job in expanding the capabilities of its gems. These are Gemini's versions of say GPTs, for example, or a Claude project. So they do have some pre-made gems that are made by Google, but if you want to create your own, you can come down here to new gem. Uh, you give it your gem a name, your instructions or your prompt basically what you want to accomplish, and then also the you can, uh, choose a default tool that it goes to each time. So, for example, if you choose deep research as your default tool. So all you'd have to do, for example, is write instructions. So for whatever topic that you put in, it just automatically goes and does research on that specific, uh, topic. And then also, of course, you can add your knowledge files down here in the field.
And as I mentioned before, the multimodal capabilities of Gemini are where it really, really shines. You can click on the tools option here. You've got access to, if you're on the pro version of Gemini, uh, create images here. That is Nanoban Pro, their image generation model, which I've talked about here on the channel. You've got access to VO3, which is their, uh, video generation model, which is also super, super cool. And then also Canvas, which I think is a really little-known feature here within Gemini that not enough people know about or use. Canvas is really good, for example, at creating presentation slides.
Now, all this said, for me, it still struggles, as I mentioned before, with content creation. Responses often feel sort of dry and lack personality, even with the really, really good Gemini 3 Pro model. And also, the lack of MCP connections is quite frustrating. It does have a few options. So, if you go down here to settings and you go to apps, you'll see a few here. So, you can connect GitHub, uh, your Google Workspace tools, which isn't surprising. You've got YouTube Music, and then three other ones here in Asana, HubSpot, and Mailchimp, but that's it.
Now, in terms of privacy, when it comes to Gemini, Gemini is going to follow Google's privacy policies. So chats, uploads, voice, screen recordings are used for training unless you opt out via Gemini apps activity, which I'll link to in the description below.
Now, in terms of limitations with the Gemini free plan, you are, there are quite a few limitations. So, you're going to have lower usage caps overall, like chats, prompts, deep research, uploads, image generation, and the context window is much smaller. So, if you're on the paid plan with Gemini, you're going to get a context window of a million tokens, which is massive. If you are on the free plan, your context window size is only going to be 32,000 tokens. So, a huge, huge difference.
So, is it worth upgrading to a paid tier when it comes to Gemini? Prior to Gemini 3 Pro, I'd have said nope, not at all. Now, I say yes. It is that good.
All right. And then next up is Perplexity. Perplexity is great for fast, real-time search with citations. So, for example, when I open up Perplexity and I want to search on best AI browsers for handling multiple tabs efficiently, I can click on go. It'll show you what it's doing in terms of thinking and then the sources that it's going to. And you can see it's really fast. Then within each of the outputs here, the results, it's going to link to the citation or the source where that specific, uh, in this case here, bullet point is coming from. It's really good for quick searches like this, including the >> [music] >> um citations.
Again, you can click on the little globe here, and you can basically tell it what type of sources that you wanted to search. So, I was searching using the web. You can turn on to include social searching for discussions and opinions. You can do, uh, SEC filings if you want to, if that is your niche. You can turn on the academic research, as I just mentioned, which is going to, which is going to research academic papers. And then if you want to include your own tools that you are using in your business like Slack or Notion or Gmail or Google Drive or what have you, you can also flip those on as well.
Then also you can choose a model. Now, this is only for the paid plan. But notice when I click on the model, I can choose any of these models that I want to use.
Now, one feature that I don't think gets talked about quite enough here in Perplexity is the tasks feature, which I really like. So, if I go under account, I can click on tasks and then I can either choose from some pre-done ones or I can click on a scheduled task, for example, and you can schedule recurring searches, for example, and have them delivered to you whenever you want.
And then Perplexity has what they call their spaces feature. They've had this for quite a while now, but this is similar to say like, uh, Chachib projects where you can organize your research into different projects. And then also, as I mentioned earlier, they've also added connectors or these are the MCP connections, very similar to say Chatebt and Claude, but way more limited. Now, I will say for the tools like Gmail, Google Drive, Notion, Slack, for example, I really like the fact that they have both read and write capabilities for the main connectors.
Now, where Perplexity really falls short, at least in my experience, in my opinion, is deep research. I don't find like the depth of the responses that I get within Perplexity's deep research anywhere near what I get in Gemini deep research or Gen Sparks deep research. And then also with Perplexity, you can do image generation in here. So you just click on account, then go to preferences, and then you'll see here under AI, you'll see image generation model. So it defaults to default, and then you can choose from here if you want say nano banana, for example, or flux.
And then finally, within this same screen here, in terms of privacy, queries and your, what your chats and conversations are logged and used to train the model here in Perplexity by default. And so unless you opt out, it will be doing that.
In terms of the free plan, what you're going to get, as you can see here, search history access, practically quote unquote unlimited basic searches, very limited amount of pro searches. Notice the ambiguity there. And you also don't have any access to the advanced AI models or image generations. Whereas with the Perplexity Pro plan, you get way more. As you can see, um, right here, you can pay either $20 a month or $200 a year for the pro plan. They also have what they call the Perplexity Max tier, which is basically for power users of Perplexity. The other thing, too, that they have for Max users is their Perplexity Assistant. Again, email assistant access, smart inbox labeling, drafting, calendar, everything in pro, etc., etc.
Now, honestly, I don't use Perplexity as much as the other tools that we're talking about here today because I'm seldom impressed with the answers that I get. So, is it worth upgrading? If you're somebody who needs or relies on up-to-date information with citations, or maybe you do a lot of academic research, I do think Perplexity is very good with their academic research, then yeah, I would say paying the $20 a month for the pro level is worth it. Otherwise, I don't think it's worth the upgrade.
So, if I could only choose one of the tools we're talking about here today, one model, a question I get all the time. Look, there's no one-size-fits-all AI tool. It depends entirely on your specific use case, like we've been talking about here today. But if I'm choosing one AI to use day in and day out and pay for that, it's going to be Claude. It's the most versatile in my opinion and consistently useful for my workflows, and also Claude code is just a whole other level. It's a game changer. But I'll save that deep dive for another video.
All right, hope this was helpful for you. Again, as I mentioned earlier, if you are an online entrepreneur and you're looking to leverage AI in your business so that you can be streamlining your business, make it more efficient, make more money in your business, and also become more profitable in the process, I want to invite you to join us inside my AI Playbook community. I'll link to it in the description below. You get a 7-day free trial. Cancel, uh, whenever you want. Again, the link's in the description below. Thanks as always for watching today's video. See you in the next one.