Speaker 1: So look, ChatGPT is great, but you only have the OpenAI models with quite a bit of tooling, but there's more out there. Like if you look here, you have all the competing models, including Claude, DeepSeek, Gemini, and more, and tools like video analysis or PowerPoint generation, and even a feature like Anthropix Artifacts, all in one interface with an option for it to automatically pick the best LLM for what you're prompting for. Now, this is a competing product to ChatGPT, but it comes at one third of the subscription price of the Teams plan. And that's why I figured it'd be a good idea to today have a look at all the features that are available here so you can get a picture of what the market has produced after two years of ChatGPT doing their own thing, and make up your mind on what fits your needs the best. But first, let me back out a little bit, because if you're a viewer of this channel, it's no secret to you that I absolutely love using ChatGPT since the very week it released. And if you follow this channel at this point, you might know that it's not me doing this by myself. I have a full team of, at this point, 11 people that help me research, test, document, and educate across everything we do. And as you might imagine, we use AI across all of it. So obviously when OpenAI came out with their Teams account, I hopped right on that and signed up the entire team. But now, after the yearly subscription running out, I actually decided to cancel it, as it doesn't really make sense for us anymore. I still think ChatGPT has some of the most useful models and toolings, but the developers within our team like using Sonnet 3.5 or O1 Pro. Both of those are not available on a Teams account. There's many other examples like this that we'll go into with this video, but the point is that if you want to get the most out of AI, I'm going to show you an alternative to ChatGPT that gives you more of these options at one third of the price of what ChatGPT Teams cost. And the site that brings all of these opportunities together at a better price is the sponsor of today's video, abacus.ai. And let me tell you, a lot of platforms like this reach out to us, and we've consistently said no to all of them just because they were lacking in some departments, whereas this one has been built so well that I feel confident in recommending it. But honestly, this video is less of a recommendation, more of an overview of all the different tools that you can expect from a tool like this, and in which departments something like ChatGPT or Enthropic might be lacking. So one, I hope to raise awareness of what the competition of ChatGPT has been up to over the past years and why sometimes you might want to prefer some of it, and two, I'll show it all to you through this platform that has all of these competing models and tools and so much more, and that's why we went with this video. And at the price of $10 per seat that you get with the Chat LLM Teams product here, it's just a good deal, and I want to say thank you to abacus.ai for sponsoring this video. But before we get into analysis of all the features and their capabilities, I really want to have a look at something that's very important here, which is the pricing. Because we'll be comparing two paid tools today, ChatGPT's team plan that comes in at $30 per month per user, versus the Chat LLM Teams plan that comes in at $10 per user, with the first month being free, but you're having to subscribe for two months at least. Significant difference, I might say. Now sure, there's other plans. You might be okay with the free plan, or you might need the pro plan, but that's really not the point of this video. We're exploring a Teams workflow, plus some of you might find this more advantageous to the plus plan. But that really depends on the features and what the platform does. So let's move on over and let's start comparing these two competing features. Okay, so let's start with model selection that you get within these two, as that is really at the heart of these. And this is not going to be an in-depth dive on each one of these models, but a bit of an overview of what my opinion is on these different models today and what the consensus seems to be in the marketplace. So if the OpenAI Teams account right here, that is running out in a few weeks, you basically have two choices that really make sense. You have GPT 4.0 and 0.1. Sure, you have the smaller models like 0.1 Mini or GPT 4.0 Mini, but generally speaking, 0.1 has just become so good compared to the preview one that these are the two that I would default to, but that's basically your choice. At this point in time, end of January 25, you have 4.0 and 0.1. Now, the model selection is something where Chat LLM really sets itself apart because it just pulls together all of these different models, including 4.0, 0.1, but then so many more like Anthropix Cloud that is still considered best in class for development. Now, some people do prefer 0.1, but it takes longer to answer, whereas when it comes to things like front-end design, Cloud Sonnet is still just the best. And a lot of people prefer Cloud Sonnet for its writing style. Still, they changed this up in 4.0, but if you want the most human-sounding model, I think Cloud Sonnet is your best bet amongst everything in the market out there. Okay, then you have the thinking model, obviously. So if you want 0.1, you don't have to give that up with this platform. And beyond that, here's a thing that I really like. I think it's been only a day since DeepSeek's R1 came onto the scene. Now, if you haven't heard about this yet, we'll talk about that in another video, but it's essentially a Chinese open-source thinking model that is open-source and 0.1 level, meaning everybody can use it, install it locally, if you have a mega AI computer, I guess. But most interestingly, it can be integrated into platforms like this. Now, OpenAI won't do this because it's a competitor to their 0.1, obviously, an open-source one, that is. But here, 24 hours since the release, ChatLLM already has access to the R1 model, so you and your entire team can just be using it in here. Beyond that, we have all the Gemini models, and these are particularly interesting if you want to do something like video upload. And here, we're skipping ahead already into some of the tools that are available here, but you can do something like video analysis, upload video files, and then it will automatically use the correct model to analyze video files for you. So you could transcribe, analyze, describe every scene within some video, all within this one interface. And beyond that, we have Grok that is rumored to soon come out with Grok Free, which is supposed to be amazing. And if you know that if Grok Free is gonna be innovating, you'll have access to it here. You have one model from Abacus themselves, and then you have some selection from some other models here. But I think that the primary ones that you should really care about, we just pointed out over here. Plus, there's an interesting thing here. Look, I created an email answering agent, something like a GPT within here that has even more advanced functionality. Now, this video is not gonna be about that, but now that we talked about the models, I do wanna point out the fact that they included this AI engineer feature within the chat LLM product, all at the same price, right? They also have this computer agent. So these two kind of go beyond the scope of this video, but let me spend like 15 seconds talking about these. So this is essentially the cloud computer use implementation where it remote controls a virtual machine that you don't have to like set up yourself externally. It's all within the product. Like you don't have to install Docker and install this on your machine. You could just go in here and tell it, show me available flights to Greece, and it will prompt itself, open up a browser, go there. Now, I've talked about this before, this product from Anthropic, this concrete product, not that capable yet, very interesting to play with, but I haven't been able to find many uses for it, but certainly a great option to have here. But the more interesting one is actually, if you go to AI engineer here, you can either create a custom chatbots, something that is very similar to GPTs in here, or you can go to create an AI agent, and this is included in the subscription, and you get this wizard, and especially these presets work super well. So if I want that email answering agent, I just go in here, I give it some custom instructions or some extra knowledge here, and literally just say submit, and then it will go through this entire process of writing all the code, hosting the code, running the code, and you can see right here, if anything goes wrong, it self-prompts itself to correct that, and the point is that here at the end, it successfully creates and deploys the simple agent that you can edit over here and use over here in brand new URLs. All I need to do is connect my Gmail account, so it has access to the emails, and now it can go ahead and craft custom draft replies to emails I might have. So it looks at the past 24 hours and it gets to work. So in this case, you can see this is an account where I received newsletters and notifications, so it skips those. It sees that, hey, if I get a security alert or some notification, there's nothing to be done here. Again, don't wanna spend too much time on this, but this is a super interesting way to learn about these autonomous workflows. During the setup, you can even tell it to be autonomous and just run regularly and do this stuff for you, but again, the focus on this video is really this competing chat GPT platform because they don't even have a product that competes with this yet, so that's more of a look into the future. Now let's talk about all these various features in here, and I wanna start with the most important one, which is this root LLM thing. This is something that we've been hoping for from OpenAI. We might get it one day, but at the end of the day, like educating yourself on what all these different models do and when to use a one and when to use 4.0, it's just something that a lot of people don't have time for, and it gets especially complex once you throw in all these others into the mix. Like when should I use DeepSeek versus Sonnet versus 4.0 versus 0.1? Root LLM will just figure it out for you in the background, and you can really get the best of all these different worlds, which is nice, especially if you're getting this for your team and they might not have the level of education that you might have, and then let's be real, things keep changing. I mean, they just added this model, figuring out what's the best every single week instead of like a full-time job. Let me tell you, it's my full-time job and it takes up all your time. Okay, so that's the first feature. The next one I wanna highlight here is this code LLM feature, and it's their own code editor that they built into the application. This goes beyond what chat GPT has. Okay, so if I run a basic prompt like create a dashboard to track weather changes in chat LLM, and then I do the same thing inside of GPT 4.0 with Canvas, this would be the feature that you would want here. Then at first, this might seem similar, right? It creates this interactive canvas for the code where you can actually make edits, which is fantastic. Then also you have the ability to follow a prompt down here, but you probably already know this. Here inside of chat LLM, this is very similar with the key difference that you can actually tab over to a preview, the little dashboard that it created. So if you're a non-coder, this is a real advantage. And I always like this feature. This is not something new. This has been available inside of Anthropics Cloud. Yet again, this feature has been implemented into this one interface that combines the best of the different apps. If you wanna go even further with development, then you might've heard that there's all of these specialized applications that do AI-powered development. They're usually a fork of an application called VS Code, and then you just have to use a completely standalone product. Chat GPT doesn't even have anything like this, but with chat LLM, yet again, they actually created their own version of this called Code LLM. This is a desktop app you download, and once you install it and go through the setup, which takes about a minute and is included in your plan, well, then you're presented with a standalone app, which is based on Microsoft Visual Studio Code. And you can do all of these features that usually you would have to get a standalone app for and pay another subscription. And again, all these different models are available within here, but the development aspect here is not the focus of the video. I just wanted to let you know that this really does combine a lot of the most popular AI applications all into one subscription. So another feature which is kind of expected at this point is the ability for you to add different files. But again, as I mentioned, because it also uses models like Gemini Pro, you can do something like upload a video and it will find the right LLM to get the job done, analyze your video file, and you can work on top of it. Of course, this also works for PDFs, markdown format, images, whatever else you might want to be doing with this. Look at that. You get a detailed breakdown of this ridiculous little clip of cats with hats at a party. This is just some test footage I generated with an AI video generator. And yeah, you can work with this stuff, all in one interface. If you try and do that inside of ChatGPT, you will quickly find that neither GPT 4.0 or 1.0 can do this. And then one thing that you can actually do here is you can go to platform connectors down here and you could connect this to various platforms. So as you can see, there's a lot of them to choose from. You could easily connect this to your Google Drive or Slack, both applications that we use here within the company. So for example, the Google Drive one you would have within ChatGPT, you can just add content directly from Google Drive with this. You can also use OneDrive here, but then connections to something like Slack where you actually activate a little chat bot, aka GPT in ChatGPT, that you built inside of this chat LLM interface directly inside of your Slack, making this yet again another feature that you don't have in here. Okay, there's just a few more things I wanna show you here. One of them would be the image and video generation capabilities. This is one of the big things you can do with AI, right? Create unique images and videos from scratch, just with text. And while ChatGPT does offer image generation, it only does that with DALI, which at this point is not even close to being a top tier model. Honestly, saying that that model is average would be an overstatement at this point. It's just not very competitive with models like Flux, Majorni, Recraft, or Ideogram. So if I go ahead and say something like generate an image of a doctor helping an injured child, and I say image, you will see that these models are available within here, and you can actually use Flux 1.1 Pro Ultra right in here. This is one of the models that you would arrive at if you looked at the entire AI image generation space. And for example, Ideogram is really good if you're gonna generate something with text or logo. And it's all here. You can just put in the prompt. Sure, I can use the prompt improver here, and I'll do a square image and say generate. Meanwhile, if I do this inside of ChatGPT, it will also do it, but we can generate images here in a second. And because this has so many models, there's just better options out there than DALI. I mean, fair enough. That's a solid picture, but this one is just better, no? Same thing goes for video generation, by the way. With ChatGPT, you do get Sora, but to really use it, you need the Pro subscription for $200. Otherwise, everything's gonna be watermarked, low resolution, and you only get a very limited amount of generations. Here, you can just go to VideoGen. And again, you get some of the best models in the market included in the subscription. At this point, this is getting kind of ridiculous. For example, KlingAI v1.6 is something that has been pointed out by a guest speaker inside of our community last week, Andrew from AIST, and he's building an AI film production company. And he was saying in their tests, Kling is just by far the best when it comes to rendering humans at this point. So yeah, you just put in the same prompt in here, and boom, you're generating a video with the best of the best that is available out there. And you have other fantastic alternatives like Runway, whereas with ChatGPT, you only get Sora. And if you wanna get rid of the watermarks, you need to pay $200. So there you go. That's the Kling generated video. Let's have a quick look. Yeah, right there. That's pretty solid. And again, I think the important point here is that you just have one account and you have everything in here from video generators, image generators, all the LLMs and more. Kind of neat. But hey, not to bash too hard on ChatGPT, right? It's a fantastic product for many reasons. And there's also some features where I would like to highlight that you won't get these in competing products. For example, the projects feature is something you don't get in this interface over here, which I personally really like. This really allows you to concentrate the conversation towards one topic. And it works even better than GPTs, in my opinion. But there is an equivalent to that inside of ChatLLM, which you can access by going up here, going to AI Engineer. And if you're creating a custom chat bot, well, this is essentially that. You can give it the goal that you're looking to achieve or the prompt that you're looking to run. And if I, for example, take my social media mentor GPT and take all of these instructions, I could either put them in a project in ChatGPT and I'll simply name the bot and give it the full instructions. And then you'll get something like a GPT that helps you organize your prompts and workflows inside of here, just like we've got the email answering agent. Yeah, there you go. Here's the social media mentor. Just like a new LLM, I can access this preset at any time. One more feature that you don't have in ChatLLM is actually if you go over to Customize ChatGPT, it's custom instructions or memories, custom instructions, more of an advanced feature that you need to know how to use. Memories just automatically gather context, which you would need to manually input here. Also the new tasks feature, which is sort of experimentative and a very early look at what the product will develop into in the future is something you don't get inside of ChatLLM. But there's specific use cases that are not in ChatGPT and it's not intuitive to use them. So if you onboard your team into one of these apps, it's for example, it's really nice to have a humanize feature here. So a prompt like this, writing an email to my boss about the broken coffee machine, I could use the humanize feature here. And then maybe I want this to be concrete to the point and helpful, or I wanna get around the eye detectors. I can put instructions in here. And then if I apply this, everything that I write inside of this chat thread will be humanized. Well, let's see what model it decides to go for. All right, so it picked GPT 4.0 and it applied this feature onto the chat in the background. So maybe I could switch over to AI detector proof and it will do some custom prompting for you in ways that are absolutely not intuitive inside of ChatGPT. So if you're onboarding your team and you don't want them to memorize all these different techniques, use cases, all these things that you can do in here from a dropdown, well then this interface is a really easy way to show them what's possible with image gen, video gen, custom prompting like this. And if you wanna add new team members, it's as easy as this. I just go up here. I say invite users and I can invite multiple team members just like this. And here I can assign people to different groups, which can then give the ability to manage our account, invite new users, add custom API keys for other services, basically giving you even more customizability than you get with ChatGPT teams. Where yes, you can equally as well invite teammates. Whereas I personally always found this team workspace quite limited. Almost everything is blocked for enterprise usage, which I believe you need something like 200 members or more to get an enterprise license. And then literally the only thing we can do here is just share the GPTs in between each other, but we could do this anyway by just publishing them. And other than that, I'm just paying more for a version of ChatGPT that doesn't live up to its original promise of delivering the newest features to the team's account first, that they never did. And on all the rollouts, now the features go to the pro users for $200, then usually to the normal and teams users. But for the team's account, they just pay 50% more just to what? Just for the ability to bulk invite people? That's why I'm making this video. I wanted to show you that, hey, if you're going to do that, then if you want to manage a team, you can use a platform like a backers.ai here and pay less, get more in many senses. And this is just a really beginner friendly way to show people all the different things that AI can do for you today. So yeah, that is the chat LLM platform. I'd be curious to hear what you think. If you want to check this out for yourself, go to the very first link in the video description below. And that's everything I have for today. See you soon.