ChatLLM: The All-in-One AI Platform in 10 Minutes

--- type: transcript date: 2025-01-23 youtube_id: hNPqEwOSaN0 --- # Transcript: ChatLLM: The All-in-One AI Platform in 10 Minutes in this video I'm going to show you Chad llm which is an all-in-one AI assistant that brings together multiple large language models and Powerful AI tools all within one place you can try this out right now by clicking the link to chat llm in the description below go ahead and check it out as we dive in first off right off the bat what is chat llm what chat llm allows you to do is access all of the best large language models such as 01 GPD 40 Gemini Sonet 3.5 all with in one place now the great thing with this is you can do this all with one single subscription you no longer have to juggle multiple tools or pay separately for each model and the great thing with chat llm is it's considerably cheaper this is an option that is only $10 a month and this is in comparison to the standard $20 a month that we see On Tools like Claude chat GPT as well as perplexity one of the things that I ran into and I'm sure you ran into last year as well was every time that there was a new feature that came out within chat GPT Claud gemini or perplexity I was definitely pulled in multiple directions trying to decide how many subscriptions am I going to have each month and honestly some months I had upwards of $100 in subscriptions to give you an idea you'll be able to access 01 this is the advanced reasoning model from open AI say if you have coding specific questions you can leverage claw 3.5 Sonet which is generally speaking the daily driver for most developers out there at least at time of recording and there are also all of the state-of-the-art open source models as well like llama 3.3 70b as well as deep seek V3 one thing that I've noticed with Abacus is every time that there is a new major model release as soon as that API is released it's within short order that they roll it within the platform that's one thing as you start to see the hype of a particular model but you might not want to subscribe to multiple different platforms this allows you to access all of these different models within one spot the nice thing with this is you can actually access some of these multimodal models for instance say if you want to leverage Gemini to be able to do some video analysis on something you might have you can do that alternatively if you want to access the latest and greatest image generation models out there they're all within here you have the flux models Dolly 3 recraft idiogram you have the ability to configure them all within here in addition to that you have the ability to search the web as well there's just a ton within here I'm definitely not going to be able to touch on all of the aspects within this video you'll definitely have to dive in and explore this yourself to really get a sense on how much this can do but even within here the state-of-the-art video generation models are in here so cling Runway AI as well as Luma Labs you can go in you can upload starting frames if you have them describe your prompt and you'll be able to generate videos the other great thing with what they've built within the application layer is you can actually have an interact with something similar to the artifacts feature that anthropic put out here we see very quickly that it generated this website for us if I say make this website about developers digest and like we will see here it will go ahead and generate that website for me here's our second iteration of that so developers digest we have articles tutorials resources newsletter I can subscribe now and overall the great thing with this is honestly the value for $10 a month is pretty wild for what you get within chat llm in addition to everything within their web app you also get access to code llm which is a fork of VSS code similar to how cursor and wind serve forked VSS code and created these powerful AI idees that are increasingly popular this is what it looks like the typical look and feel of vs code and you'll have the panel that you can toggle back and forth for code llm within here you can choose from a number of different models including 01 as well as One Mini this is a pretty big deal cuz within here with just one subscription I can access these state-ofthe-art models this is a new offering and I'm particularly excited with this I generally spend most of my day within my IDE and being able to access both 01 01 mini as well as Claude 3.5 Sonet which is what I typically use within my editor is pretty powerful I'm not aware of any other company on the market that charges 10 bucks for their IDE their chat interface that also has a mobile app and all of these other features in addition to that there are a number of helpful features that are built right within here you can humanize text you can also open up an editor if you want to take a screenshot of something that you might have to demonstrate video analysis for instance I can put in a URL to even a YouTube video and I can say tell me why this video performed well I'll submit that and then we see it's going to go it's going to send in that video this basically has a use case regardless of what discipl you work in whether you're a developer there's things for you whether you work in other aspects of other Industries there will definitely be something within here that you could leverage here we go it analyzed the video here it has all the different details of the video the author the title the URL as well as all of the details and within here it gives me a summary it's clear concise it's timely relevant it has strong educational value practical tips insights so on and so forth another great feature is you can actually index your GitHub repositories as well what I can do here is I can just select the different folders that I want to index I can go ahead and add those files after a few seconds after it has all been indexed what you can do is I can just ask how do I add in a new conditionally rendered UI component and shortly thereafter it's giving us good instructions based on a relatively large GitHub repository that I have it's giving me exact instructions on how to add in a new conditionally rendered UI component and then in addition what you can do is if you want to try and make changes to a GitHub repository you can go ahead and actually try and attempt to make a pull request so this isn't something that I haven't tried out too much but what's great with this platform is there is just so much to explore another feature within the platform is you can leverage their computer agent this is similar to the cloud computer use feature that was released just a number of months ago where you could spin up an instance and what it would allow you to do is with natural language effectively control your computer here's a demonstra where I'm asking to find and download an image of matchu Picchu what it did there it opened up Firefox we can see that it's typing into images.google.com and it has the full control of being able to use a cursor as well as type in the keyboard here you see that within Google Images it went and it searched for matchu picu and it has all the natural language responses on the left hand side of exactly what it's doing it's clicking the first image it's going through and it's actually going and clicking to save the image overall there are just a ton of features within here you can even scrape URL so if I go to Sam alman's blog and if I just say summarize the key points of the latest blog I'll submit that and very quickly it's going to go to that website it's going to scrape the website it's going to feed it within the selected llm that you have and it will analyze the results here based on the latest blog post titled Reflections from Sam alman's blog here are the key points we go through this and there are a ton of relevant details future outlook states that open AI is confident we know how to build AGI predicts that AI agents quote unquote join the platform and significantly impact companies and so on and so forth overall the platform is really just an aggregate of everything that's happening within AI from image generation models through to the latest and greatest Frontier models all the way through to application layer features like artifacts being able to generate PowerPoints being able to create documents and like I mentioned even having access to their new code editor last what I want to show you is their AI engineer feature within here if you ask for something like create an AI agent it will say I'll help you create an AI agent let me know some sample AI agents to get started you can either choose from one of these samples or create your own agent here let's say I want an email answering agent do we want it to be autonomous I can say no do we want reference information I can upload that and within here we can add in additional instructions like maintain a professional tone and undersign the replies with regards develop digust also subit that and here we go it's Gathering the code to define the AI agent in just a number of seconds we have this relatively large block of python that we could go and use for executing a new AI agent here we even have examples and what the execution results will be if we just go through we can see exactly what this looks like what's really cool with this is what it will do is it will Define the code for us and once the code has been defined we're going going to have the execution result we see that the execution completed successfully it's registering the AI agent I'll help you create an email answering agent as mentioned in the specification this agent would be difficult to dry run locally because it needs actual Gmail addresses and unread emails to work I'll skip through the local testing and directly create the agent here we go it's executing the results we see the execution results we see the errors and what it's doing is it's just going to pass in those errors and it's going to keep going through the process until I created this agent with this ID look at unread emails in the Gmail box it's deploying the code and now I can go to the app I can log into Gmail and I'd have this link that I can access where I can log into my Gmail account to have access to this agent this is really neat because a few different clicks it went through this flow where it was able to write the code execute the code test the code and actually even deploy the app for us that's pretty much it for this video as you can see you get an absolute ton of different features with in the platform when you sign up for an account from the latest Frontier models to the ability to generate images all the way through to creating and deploying your own AI agents I encourage you to check out chat llm and try it out for yourself if you found this video useful please like comment share and subscribe otherwise until the next one