NEW DeepSeek R1 & Bolt.DIY!🤖 Qwen 2.5 Max + Groq R1 Llama Distill Open Source FREE AI Coding Agent

Everyone has been talking about deep seek and their crazy new good reasoning model R1 and I've recently been doing a lot of different videos showcasing how you can use deep seek R1 in different use cases like an n8n or en cursor or with Gro and I even just did a video on quen 2.5 Max which showcases how China is really taking a huge step forward in terms of the AI race and open source AI now with quen 2.5 Max they did

Comparisons in their benchmarks between deep seek V3 not R1 but I think it would be cool if quen AKA Alibaba did come up with a reasoning model themselves now there's a few different ways you can go ahead and use these models of course you can use the cloud version through their chat which is useful if you just want to go ahead and test these models out but what's some of the best use cases you can use if you want to go

Ahead and actually build an app or use this for AI coding well in today's video we're going to go over one of those use cases and that is with bolt. DIY now I've done a few videos in the past covering bolt. DIY I covered it with deep seek version 3 I covered it before when it was called autod Dev and if you're not familiar with bolt. DIY it is an open-source version of bolt. new and it was originally created by Cole

Medan and his community now this is a really good tool because it allows us to test multiple different llms and test it in a self-hosted containerized coding interface where we can see which llm performs best so I'm going to show you how you can get this set up self-hosted in the next 5 minutes how you can use it with deep seek R1 to build really cool apps as well as quen 2.5 Max and also some potential different ways you may want

To use deep seek R1 without Ado let's dive right into it so like I mentioned deep See's been generating so much hype in the AI industry and the AI space and quite frankly the mainstream media too I know people that have nothing to do with AI that know about deep seek now and it was number one on the app store for their app you can see it's caused open AI to release different statements like Sam mman doing different tweets like saying

The model is impressive but we will obviously deliver much better models there's been a couple open AI employees that are salty and tweeting out as well about deep seek and as you can see open AI says it has evidence that China's deep seek used its model to train competitor the San Francisco base chat GPT maker told Financial Times it has seen some evidence of distillation which it suspects to from Deep seek the technique is used by a developer to obtain better

Performance on smaller models by using outputs from larger more Capital ones allowing them to achieve similar results on specific tasks at a much lower cost distillation is a common practice in the industry but the concern was that deep seek may be using it to build its own rival model which is a breach of open AI terms of service I mean I wouldn't be surprised I mean deep seek did distill quen as well as llama in some of their model releases and

There's also been some conspiracies such as nvidia's back door to China saying that recent financial data and market analysis have raised questions about nvidia's sales patterns and potential circumvention of us export restrictions while nid reports that direct shipments to Singapore are insignificant 22% of last quarter's Billings were to Singapore addresses fueling speculation about a Singapore back door or chip sales to restricted markets like China but enough with the conspiracies let me know in the comments down below what you guys think

So far about deep seek and are some of these conspiracies actually true but let's go ahead and get get started with bolt. DIY so first things first if you don't already have bolt. DIY self-hosted you're going to go over copy the GitHub link that is in the description down below and we're going to go over to our terminal right here and we're simply going to run get clone and then the repo link right here once you run that you're going to

Change directory into bolt. DIY and then you're going to open it up in cursor or vs code so you could run code dot now once you have this open you're going to have a bunch of different options in terms of API keys that you can connect so you're going to see a env. example you're going to want to copy this and duplicate it and rename it Tov and then you can really just start pasting in your API key so what I

Would suggest for this is Gro because they have deep seek distilled which you'll get Lightning Fast inference speed so you can go ahead test this out you could also use hugging face open AI anthropic open router you know you can access deep seek through open router to Google generative AI olama so you can also use ol llama you can put in the API base URL here and as you can see you can get deep seek R1 on oama as well now

You probably won't be able to use some of these bigger models unless you have a really powerful GPU but you could go ahead and use maybe the 14b model 8B model maybe even 32b depending on your hardware specs now you can also use an open AI like API base URL so if you want to use quen 2.5 Max then you would want to go ahead and paste that open AI like API based URL right here so that would be- scope Das

Intel this right here now I'll leave a link to that video I did yesterday if you do want to use quen 2.5 Max you can go ahead check that out are you tired of pouring thousands of dollars into appointment Setters only to watch leads slip away imagine having a team of elite sales agents booking qualified appointments for you around the clock no more wasted time on training no more frustration with performance and no more draining your budget on inconsistent and expensive

Call centers introducing stride agents AI powered appointment Setters that work 24/7 never get tired and book appointments while you sleep trained on thousands of successful conversations our AI agents outperform human teams at just one tenth of the cost join the ranks of businesses that doubled their appointments and booking rates in just a matter of weeks don't get left behind in the AI Revolution visit stride agents.com now and transform your entire sales process with cuttingedge AI technology it's time to accelerate your

Stride with AI agents you would have to create an API account with Alibaba so if you don't care about China having your data then you can go ahead and do that now we also have some other platforms that do have deep seek like together AI which does provide you some free credits we also have deep seek API key so this is the direct API key from Deep seek now currently if you try to go to platform. deepsec it says our website

Is currently under maintenance we apologize for the inconvenience now this significant disruption is apparently due to a large scale malicious tack that began affecting its services January 27th so this forced them to temporarily restrict new user registrations apparently this was a Dos attack targeting the company's API and web chat platform so who knows maybe I'm not saying any names maybe Sam alman's pulling the trigger I don't know we did hear some conspiracies with the whole open AI whistleblower so who knows

Not saying I actually believe that but if you do let me know in the comments down below I'd love to hear some of your guys' conspiracy theories but if you do have your deep seek API key handy would want to paste that in there now one thing I will note too is you can paste in all your API keys right here before you actually do anything or once we actually get the app up you can actually paste in your API Keys

Within the UI but if you want them permanently there you can paste them in right here in the environment variables all right so hyperbolic also has deep seek available so you can check that out as well and I believe they have some free credits we also got other platforms like mraw coh here LM studi so you could use LM Studio instead of ama if if you're running deep seek locally and then we have xai perplexity Etc so put in the ones

You want you know the specific ones to deep seek that I outlined and then what we're going to do is we're simply going to run pmpm install to install the dependencies now one thing I will say is if you already have bolt. DIY downloaded on your system then you can simply just use get pull origin main to pull all the recent updates since they implemented deep seek Integrations now once you run PMP install and the dependencies are installed You're simply just

Going to run PMP run Dev to start the development server now if you have any issues with this you can check out the GitHub repo right here which walks you through step by step as well as different ways to install it using Docker Etc but now this is going to start bolt. DIY on Local Host 5173 now first things first if you want to use quen 2.5 Max you can go to the open AI like section right here and if you

Paste it in that base URL like I showed you then you can simply just add in the open AI API key right here for Alibaba or you could do that in the backend too and that's going to pull those models so we have quen Max quen Plus quen turbo instruct Etc so I'm specifically going to focus on deep seek right now because deep seek R1 is just awesome and it's even cheaper so let's go ahead to deep seek now within deep

Seek we can see deep seek chat deep seek coder and boom deep seek Reasoner and this is obviously the one that has been shaking up the AI industry that we're going to use now like I mentioned you could also use other platforms like hyperbolic here or together but we're just going to use the direct API all right so I'm going to say build a beautiful CRM application that is modern and has a Shaden like style impress me I even say make

It on a react framework and let's go ahead and send this and like I mentioned you could add API keys right here too if you didn't add them in the EnV file and it even gives you a place right here to get API key which will take you directly to the API section now this is really cool because they even integrated a thought process already so you can see deep seek thinking within bolt. DIY so this is really crazy so I

Really like reading this even though sometimes it's long I'm not going to read the whole thing but this gives you an insight to how a really powerful in intelligent model thinks through different problems this is really cool because it makes the whole process very transparent opposed to open AI 01 model where they hide the thinking okay we're looks like we're getting an issue here this operation is insecure I don't know why we're getting this issue just spammed right here but it

Looks like it is creating the app so let's go ahead and read some of this so we can see all right the user wants to build a beautiful CRM application with a modern style similar to Shaden built using react let me first break down how to approach this first I need to set up the project structure using V with react and typescript make sense for a fast setup I'll include Tailwind CSS for styling since shaten is built on it but since

Third party libraries like shaten can be installed VI by a pip I'll have to rely on mpm packages that are compatible with web containers limitations next the UI component Shaden style is clean and modern so I'll create reusable components like button card data table using Twin's utility classes will help achieve that look without external dependencies I'll need to set up the theme configuration in Tailwind config to include custom colors and fonts and you can see it's continuing on with its reasoning

Which is really cool to actually read and as you can see it is building out the files right here we got we got the package.json we got the main. TSX index. TSX app. TSX we got our different components like dashboard button card data table drop- down menu and boom it is running the command npm install Okay so boom We encountered an error while running the terminal commands would you like bolt to analyze this let's go ahead and ask bolt we can

Click through see the process okay the users getting this error right here it looks like looking at the package.json I can see that that V is listed so had to do a bit of troubleshooting but there we have it we got our CRM dashboard contacts deals analytics as you can see we could export new contact now of course you could build upon this this is really just the foundations right here and like I mentioned before we could go back to a

New chat and this is of course where you could use like I said o Lama or open a I like or any one of the other Alternatives but just like I showed you in my last video we're going to use Gro right here and we're going to use the Deep seek distilled R1 llama 70b which is going to be blazing fast with a very powerful model I'm just going to say build me a to-do app and react using tailwind and we

Will see how quickly it will actually go ahead and start doing this so we can see the thinking process right here okay and look how quickly it is actually getting to work on these files it's actually blazing fast and it's already done it's running npm install I don't know why it keeps doing this operation is insecure but anywh who it is launching the app as we speak I think this is extremely powerful because when you take something like Gro which is

Very quick inference time and you put it something with something like DEC R1 which is a reasoning model that you know has to think and it takes time to think but if you could speed up that thinking process and get that same quality that's very powerful so once Gro gets some of the larger models that's going to be insane as well all right now we are getting a rate limit error right here just cuz we did have an issue I was

Trying to troubleshoot it but we're getting rate limited by grock which is fine I think you get the point so that's pretty much it for this video guys on how to use deep seek R1 which is an extremely powerful open- Source model in bolt. DIY which is an extremely powerful open-source AI coding tool so if you're new to the channel we upload videos all the time on AI Marketing sales business growth so if you want to stay out today on the

Latest AI Cutting Edge technology open source tools make sure to subscribe like the video and comment down below let me know what your thoughts are about deep seek R1 all this controversy all this drama China is literally taking the open source a industry by storm right now and the whole world is watching to see what the USA will do about it also too guys if you haven't already joined our free Facebook group and Discord Channel strid community.com I'll leave a link

Down below you're definitely going to want to check that out and then also too guys check out our stride AI Academy for behind the scenes resources templates tools trainings Etc it's going to be free at the moment but it may not be free in the future so join while you still can also too guys if you run a business and you need help implementing ai ai agents AI appointment cers into your business then call down below at executive.com apply and we

Can see if it's a fit or not other than that guys I will see you in the next video keep hustling keep grinding and of course guys accelerate your stride take care