NEW China AI Qwen 2.5 Max Beats DeepSeek v3!?🤖 Open Source FREE LLM AI By Alibaba (n8n AI Agents)

So lately everyone in the AI industry has been talking about deep seek R1 and quite frankly everyone just around the world in the mainstream media and everywhere has been talking about deep seek R1 because it is completely changing the game when it comes to AI open source AI as well as China versus Us in the AI race now I've made a few videos on deep seek in the last couple days deep seek R1 how to use it through the API and

Tools like cursor n8x use it with Gro as many different providers are coming out with deep seek right now and a lot of people are saying oh my God China is really taking a a big leap here especially with open source now I definitely wouldn't count USA or open AI or whoever out of the race by any means and you know we haven't seen at all what you know some of these companies have behind the scenes so it will be really

Interesting to see what is up next but deep seek is still going they recently just dropped another open source AI model called Janice Pro 7B and it's a multimodal model that can generate images and apparently beats open AI dially 3 stable diffusion gen eval and and DPG bench benchmarks what's also crazy is deep seek R1 just got a two times speed boost and guess what the Boost was written with deep seek R1 itself so it's already self-improving so self-improving AI is

Here we're really going to see more of that coming soon and it's not a matter of time before we get into that reflexive cycle where we completely remove human in the loop and it's going to get crazy but deep seek is not the only Chinese company that is pumping out AI updates quen which is a model owned by the company Alibaba just released their quen 2.5 Max version and the benchmarks are quite impressive as you can see quen 2.5 Max is

Beating deep seek D3 which is the model that deep seek released before their reasoning model on many different benchmarks across the board as you can see here and it's even beating models like Claude Sonet 3.5 GPT 40 llama 3.1 now like I've always said on this channel benchmarks are definitely not everything and it always comes down to actually testing it because you can't just always trust benchmarks especially when the company rating the model is the one putting out these Benchmark tests

But that's why in today's video we are going to run some tests with quen 2.5 Max alibaba's newest model I'm going to show you how you can use this open- Source model 100% for free now you'll be able to use this either in their Cloud version or in their API and I'm even going to show you how you can connect it to tools like n8n so if you do want to use quen 2.5 Max and an n8n AI agent you can

Definitely do that so without Ado let's Dive Right In now all links I cover in today's video will be linked down below as you can see quen also released 2.5 Vision so there's different parameter models right here like 3B 7B 72b and if you want you can check out their blog Down Below on their Vision language model and you can see a comparison between benchmarks between some of the models like Gemini 2 flash GPT 40 CL 3.5 quen 2 and they

Also release quen 2.5 1 M so deploy your own quen with contacts up to 1 million tokens and then they go on talking about long contact Tas versus short context task like I said all links will be down below if you want to check more of that out but we're going to focus on the 2.5 Max right now so there's a few ways you can use it like I said quen chat the API or the demo so the demo is here

Within hugging face you go ahead test this out if you want I'm just going to use their actual chat right here now you'll see their interface is similar to really any chat model whether that's chat GPT or deep seek you can see that we have a few different toggles here which is nice so we can actually toggle on artifacts we can toggle on web search image generation right here you could select the different sizes right there you could even do video

Generation which is pretty cool now you can't have web search or artifacts toggle on at the same time and just as a side note you can upload a document or attach an image if the model supports it so I'm going to go ahead and say code a beautiful modern website for an Ecom beauty store with quen 2.5 Max and let's go ahead and send this okay and you can see it is generating the HTML right here now while this is loading

Just a couple more tweets I seen on quen 2.5 we can see the VL version right here can perform computer use out of the box taking on open AI operator headon as you can see it's being used right here with computer use Quin Max is giving out Red Envelope slopes I'm incredibly powerful search great [Music] coding wo classic strawberry puzzle [Music] game let's play the game [Music] game over prosperous New Year may your life be filled with joy just this of

Course not see you okay so let's go ahead and take a look at our Ecom beauty store and boom voila we got a you know a beauty store here obviously it's not that great out of box but you know it did succeed with the task here as as you can see the artifact right here is pretty nice within their actual Cloud version now we can also use the web search so how does deep seek V3 compared to quen 2.5 Max and

Boom as you can see we got our search results right here for the sources and you can see that it's saying that you know it has been reported to help perform dcp3 and several benchmarks now would be really interesting if quen dropped a reasoning model themselves I think that would be really cool because if you've been following this channel for a while we've actually done many different videos on quen and deep seek we've done comparison videos in the past months ago

As you can see we did a video on quen 2.5 32b here another video on quen 2.5 4 months ago deep seek 2.5 right here and these two open-source models and companies from China have always kind of been somewhat Knack and Knack so it will be very interesting if they do drop a reasoning model now the one negative with quen 2.5 Max is they haven't dropped the weights I've heard people say they typically don't drop the max weights for the model

But if you want to actually use this through the API and use this in something like maybe n to use with an agent you can go ahead to this article and I'll also leave the direct link down below and click on API now this is going to take you to Alibaba cloud and you're going to go ahead and create an account this is of course if you don't mind creating an account with a Chinese company now once you create an account

They're going to ask you to verify your phone number and you have to add your credit card as well so it's not the easiest AI API to get started with but if you decide to do that you can go ahead directly to this link I'll leave it in the description down below and then once you have your account created just click on quen on model Studio once you do that it will take you to the model studio and you'll be able

To see all the different models right here there's plug-in galleries right here with calculator python code interpreter image generation you can create applications right here there's application components like prompt engineering custom plugins data management data application so pretty cool but we're going to go to the model section right here we're going to click on quen Max now we're actually going to go to the API example and you're going to see we got an open AI base URL right here we're going

To actually copy this base URL we're going to go over to n8n and we're actually just going to create an AI agent so you can simply go up here create an agent and you're going to add a chat model you're going to go down here to open Ai and you're going to add a new connection by going here create new credentials within these credentials you're going to paste in that base URL right here then you're going to go back and you're

Going to click up here and go to API Keys once you do that create an API key click on view copy that API key and then paste it right here click on Save and then you'll be connected now also leave the pricing right here as you can see we got uh limited time free for some of the models like the VL and the VL Max and the embedding uh quen turbo right here we have price per thousand tokens quen plus right

Here and quen Max we got 0.01 cents per input and 0.03 cents per output so very very cheap and if you sign up at least right now they gifted you uh beginners 1 million free inference tokens for each model up to 50 million tokens free valid for 180 days so I did get that bonus now if we go ahead and test this model saying hey hey what model is this exactly is this quen 2.5 Max as you can see it said

It's quen from Alibaba and it says yes I'm the latest version quen Max now one thing I forgot to mention is once you actually make that connection you're going to want to go ahead and click on expression over here in the model it's going to be on fixed you don't want to use that for a drop down and you're simply going to want to come back to the model copy this model ID right here and paste that bad boy in all

Right guys so that's pretty much it for this video on quen 2.5 Max how you can use it right now in their Cloud version or in the AP Pi how you could even use it within n to build AI agents and a brief overview of some of the Benchmark comparisons between deep seek 3 and quen 2.5 Max so like I said China is really on a rule it will be interesting to see how the US fires back with their new AI

Developments let me know in the comments down below what you guys think about quen what you think about deep seek deep seek R1 deeps V3 what does all this mean for the AI Community for the open-source Community other that guys if you're new to this channel we upload videos on the time on AI Marketing sales business growth so if you like that type of content you got some value here make sure to like the video comment down below and subscribe to

Stay up to date with the uploads also too guys if you haven't already joined our free Facebook group and Discord Channel I'll leave a link down below stri community.com and then also too guys if you haven't already joined our stride AI Academy definitely check that out while it is still free also too guys if you run a business and need help implementing AI agents like AI appointment cers Ai call center or optimizing your systems your CR M your marketing whatever the

Case may be then book a call down below at executiv strive.com slapp and we can see if it's a fit or not to work together 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