NEW DeepClaude Beats o1 & Deepseek R1!?🤖 R1 + Claude Sonnet 3.5 Open Source API Self Hosted

What's going on guys now on this channel we've been talking about deep seek R1 as well as other open-source models from China such as quen 2.5 Max and everyone's really been talking about China versus USA in this AI war and even the mainstream media has been talking about deep seek in fact some people are even saying why are we allowing American Tech giants like Microsoft to help China develop deep seek and other AI it's insane we should ban the transfer of

AI technology to China immediately in fact there's even a new bill in the US Congress called the decoupling of America's artificial intelligent capabilities from China Act of 2025 that would ban the import of any AI technology from China including open source models like deep seek it would make it illegal to do any research or development in Ai and collaboration with an entity of concern defined as any Chinese institution or company now I didn't go through the bill and read it myself

Nor do I know the likelihood of it actually getting passed but this post says a new bill is being proposed to make d downloading deep seek in the US a crime with penalties up to 20 years in prison which if true and if it did get passed would absolutely be insane and like I mentioned I showed you how to use deep SE garan with ruk Cline I've showed you how to use it with wind surf bolt. DIY uh n8n cursor many

Different ways but what if you want to use deep seek R1 with something maybe like Claude Sonet 3.5 because you know many of you guys love claw 3.5 as well as me because it's so good with creativity and coding and it's just still one of the best if not the best Frontier Model well now you can with an open-source free tool called Deep CLA so deep CLA combines CLA Sonet 3.5 with deep seek R1 Chain of Thought reasoning to outperform open

Ai1 deep seek R1 and Claud Sonet 3.5 100% free and open source so I'm going to show you how you can use their Cloud version for free of course you will have to bring your own deep seek and clawset API keys but I'm also going to show you how you can self Host this on your system and then you could actually use this to make API calls and do some really cool things with this so without Ado let's dive right into

It all right guys so all links I cover in today's video will be linked down below I'll even leave a link to the actual uh build that I was talking about earlier there but let's take a look at the Deep claw GitHub here so harness the power of deep seek r1's reasoning and claw's creativity and code generation capabilities with a unified API and chat interface so this is really cool and this got me excited when I found this so deep Claw

Is a high performing llm inference API that combines deep seek r1's Chain of Thought reasoning capabilities with anthropics creative code and generation prowess it provides a unified inter interface for leveraging the strengths of both models while maintaining complete control over your API keys and data now just to caveat if you are using the Deep seek API which goes through China the hosted version then you know you may not be 100% in control of that data so a few features we have

Zero latency instant responses with r1's Chain of Thought followed by claud's response in a single stream powered by high performance rust API and then private and secure so endtoend security with local API key man management your data stays private and then highly configurable dual AI power and then manage bring your own Keys API so y R1 plus claw so deep seeks Chain of Thought Trace demonstrates deep reasoning to the point of an llm experiencing metacognition correcting itself thinking about edge cases

And Performing quasi Monti Carlo tree search in natural language however R1 lacks in cogeneration creativity and conversation skills and I can definitely say that I personally agree because even still I will use Clawson at 3.5 for many different things because it's just so good and that's where it excels in these areas making it the perfect compliment so deep Claud combines both models to provide r1's exceptional reasoning and problem solving capabilities claud's Superior codenation and creativity fast streaming responses in a single

API and complete control over your API and they even provided some Benchmark tests right here so R1 plus Sonet set SOA on AER polyot Benchmark you can see here in this benchmark test R1 plus Sonet o1 R1 Sonet and then deep seek V3 and R1 plus Sonic is at the very top and this is great because quad anthropic has not dropped a reasoning model yet so this is kind of like a cheat code for that until they do so according to

Aer's polyot benchmarks the combination of R1 as architect and Sonet as editor has achieved new SOA of 64% on AER polyot Benchmark 14 times cost reduction compared to the previous soda results improved accuracy across multi-programming languages and the beauty behind this is you can use this in their hosted version and chat with it and test it out which I'll show you right now but I'm also going to show you how you can actually self-host this and then you can use it

In an API in the application of some sort or whatever you want and this was created by a company called Aster but let's go ahead and check out their hosted version right now so like I said this will be linked down below to deep claw.com once you go to deep claw.com you're going to go ahead and click on configure up here and you'll see that you have your deep seek API token your anthropic API token and then your system prompt right

Here and then deep SE configuration if you did want to add some extra things like headers body and then anthropic configuration down here which allows you to add different things like headers right here more headers or body as well so grab your API tokens plug those in click on Save and then you're good to go so now we can go ahead and click on type message where we can compose our prompt I'm just going to say how many RS are in

Strawberry so very basic to start off we're going to go ahead and click on send 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 qualify 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 centers 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 now we can see that our assistant is thinking we can

See it's thinking for 10 seconds right here and if we open up this uh drop down we can see that we have our deep seek thinking right here so this is the whole deep seek reasoning and it's going through starting with s then t r a w b e r r y counting them back and forth making sure that it's correct and now we're getting our result this is generated by Claud Sonet 3.5 so there are three Rs in the word

Strawberry let's break it down blah blah blah blah blah there you go now you can also switch between different models like three Opus uh 3.5 hi cou we're going to go ahead and keep it on Sonet latest right here all right next I'm going to say build me the game of snake in a python file and tell me how to run it we're going to go ahead and click on send all right so we can see it is stinking right here

Going over the different libraries it needs to use all that good stuff and now we're getting our generation from Claude so here's the complete code use it with py game right here so I'm going to go ahead copy this code and paste it over into a python file right here all right now I'm going to run pip installed py game uh which I already have so I'm actually just going to run python snake game right here and boom now we have

Our snake game okay so I'm playing the snake game let's get the food okay we got the food we have a scoreboard up there everything's working really well see if I hit the side okay the side then it closes so boom great job all right so I think you guys get the point now let's go ahead and actually show you how you can self Host this so you could use this in applications you want whatever the case may be so first

Things first you're going to get clone the repo so you're going to open up your terminal and you're going to run get clone the repo link right here then you're going to change directory into deep CLA and you're going to run cargo build-- relase this is on Rust so if you don't have rust make sure you have that downloaded you're also going to need strawberry Pearl um so download this je I'll leave a link down below if you don't already have

It and if you encounter any issues when running cargo release like I did here initially then just simply copy this right here paste it into chat GPT and ask it what you need to do once you have that done it may take a little bit as you can see mine took 13 minutes but then it actually was good to go so then I ran code dot to open it in VSS code and you'll see that there's a Docker file right here

So you could use Docker but we're going to go ahead and actually just use the executable file so then the Deep CLA repo you're going to go ahead and click on Target and then you're going to click on release and then you'll see deep claw right here now if you want to change where the port is open or any of the configurations what you're going to do is copy this config file right here config dotl and paste it in the root

Of the release right here so you can see I have that here you could go ahead and change the port if you want or anything like this but I'm just going to keep it the same now if you don't do this config step that's okay it's actually just going to use the default Port which is Port 3000 so once you do that you can simply open up deep claw application right here and it's going to open up this terminal right here

And if I go to Port 3000 you're actually just going to see a blank screen like this so it's not giving you a web UI or interface but this is where you're actually able to make API calls to it so if I go to deep claw.com and then click on API docs this is where you actually see the base URL so the cool thing is you can actually use their manage API or your self-hosted version so if I copy this right

Here go over to postman right here and paste this in and then click Send and boom you can see that we got our response right here so I'm going to show you what this actually is in just a second but you can use like I said their API base URL right here or your hosted one which is your host so for us it would just be Local Host and then Port 3000 now you could actually deploy this in a server too

If you wanted and use that now for authentication you have deep seek API token and X anthropic API token so you're going to pass those in through headers so this is what your request is going to look like you want to pass in content type application Jon on X deep seek API token and then your API token and then X anthropic API token and then that API token those are the headers right there and then for the body you simply want

To pass in the message the rule and the content right here so the content we have how many RS are in the word strawberry and you can see we got our response right here so created content text so you can see the thinking right here it's thinking thinking thinking and then the text there are three Rs in the word strawberry we can see the cost right here so the total cost the reasoning cost cash all that good stuff all the different

Metrics you would want and you'll see in the API docs all the different responses error handling common error types uh non-streaming API usage as well as streaming API usage so really cool that they have this API right here and you could do some really cool stuff with this you can integrate this in your applications and I think this is actually really useful just by looking look at these benchmarks right here you can have uh R1 as the architect basically and Sonet

As the model that is actually doing the output so other than that guys that's pretty much it for this video hope you got some value here if you did and you're new to the channel make sure to like the video comment down below and subscribe to stay up to date with the uploads we upload videos all the time on ai ai coding AI agents so if you like that type of content you're in the right place also too guys if you

Haven't already joined our free Facebook group and Discord Channel stri community.com I'll leave a link down below and definitely check out our stride AI Academy it's currently free right now but it may not be in the future so join while you still can also too guys if you run a business and you need help implementing AI agents like AI appointment cers Ai call center literally to pack your calendar with with boat loads of qualified leads then book a 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 let me know what your thoughts are about deep seek R1 as well as Claude and deep Claude let me know what you guys think of this and as always keep hustling keep grinding and of course guys accelerate your stride take care