ChatGPT Vs DeepSeek VS Qwen 2.5 Max: Who Wins?
12:46

ChatGPT Vs DeepSeek VS Qwen 2.5 Max: Who Wins?

Julian Goldie SEO 30.01.2025 45 087 просмотров 786 лайков обн. 18.02.2026
Поделиться Telegram VK Бот
Транскрипт Скачать .md
Анализ с AI
Описание видео
Want to get more customers, make more profit & save 100s of hours with AI? Join me in the AI Profit Boardroom: https://go.juliangoldie.com/ai-profit-boardroom 🚀 Get a FREE SEO strategy Session + Discount Now: https://go.juliangoldie.com/strategy-session - PLUS get 25% extra FREE on long term contracts! Want More Leads, Traffic & Sales with AI? 🚀 Automate your marketing, scale your business, and save 100s of hours with AI! 👉 https://www.skool.com/ai-profit-lab-7462/about 🤯 Want more money, traffic and sales from SEO? Join the SEO Elite Circle👇 https://go.juliangoldie.com/register Click below for FREE access to ✅ 50 FREE AI SEO TOOLS 🔥 200+ AI SEO Prompts! 📈 FREE AI SEO COMMUNITY with 2,000 SEOs ! 🚀 Free AI SEO Course 🏆 Plus TODAY's Video NOTES... https://go.juliangoldie.com/chat-gpt-prompts - Join our FREE AI SEO Accelerator here: https://www.facebook.com/groups/aiseomastermind - Need consulting? Book a call with us here: https://link.juliangoldie.com/widget/bookings/seo-gameplanesov12 DeepSeek-R1 VS ChatGPT VS Qwen 2.5 Max: Who Wins? DeepSeek R1 vs Qwen 2.5 Max vs ChatGPT: Comprehensive AI Model Comparison In this video, we compare three AI models: DeepSeek R1, Qwen 2.5 Max, and ChatGPT, to determine which one performs best in various tasks. We test capabilities like coding a snake game in HTML, real-time web search, image generation, video generation, and the ability to run locally. Qwen 2.5 Max excels in speed and accuracy, especially in coding games and real-time search. DeepSeek R1, while slower and less accurate, offers the advantage of running locally and providing a free API. ChatGPT shows strong performance in video generation but lacks some features unless upgraded to ChatGPT-4.0. Watch the full comparison to discover which AI model suits your needs best! 00:00 Introduction to AI Model Comparison 00:13 HTML Snake Game Test 01:45 Web Search Feature Test 03:31 Image Generation Showdown 06:29 Video Generation Capabilities 08:40 Local Hosting and API Access 11:56 Conclusion and Recommendations

Оглавление (7 сегментов)

  1. 0:00 Introduction to AI Model Comparison 41 сл.
  2. 0:13 HTML Snake Game Test 327 сл.
  3. 1:45 Web Search Feature Test 341 сл.
  4. 3:31 Image Generation Showdown 604 сл.
  5. 6:29 Video Generation Capabilities 438 сл.
  6. 8:40 Local Hosting and API Access 682 сл.
  7. 11:56 Conclusion and Recommendations 180 сл.
0:00

Introduction to AI Model Comparison

today we're going to be comparing deep seek R1 versus quen 2. 5 Max which is the new version just released from Alibaba this week versus chat G pt4 and we going to see which is the best model now so
0:13

HTML Snake Game Test

the first test that we're going to go with is we're going to make sure we have deep seek R1 and we're going to compare how each performs when we say build a snake game in HTML so we're going to hit into like do the same over here and select artifacts which is a feature for building tools and then we're going to do this in inside check gp01 mini now R1 is probably the slowest model by far from what I've seen now in terms of speed chat gpt1 is absolutely flying through it's already created the output deep seek out is still coding out let's have a look at quen is just loading the artifact right here but it's not come up yet if you want something fast chat GPT is already winning now let's preview the outputs we're going to click on edit over here and then preview and we can actually preview the code inside chat now if we look at this game not very good now if you actually look at this game we actually can't play it I can't move around I can't direct anything and there's not much of a snake not a great start let's test deep seek R1 so we're going to click on run HTML see what we get back we'll click on start game and that just put game over straight away and now inside quen 2. 5 is actually working look at that perfect so in terms of outputs quen 2. 5 Max even though it's not a logic game it's still smashing it wow bear in mind chat gpt1 and deep seek R1 are supposed to be reasoning models so they're supposed to think these things out they're supposed to be better for logic but quen 2. 5 Max is absolutely smashing it look at that perfect much better than what we got back here that's blowing me away honestly next up what
1:45

Web Search Feature Test

we're going to be testing is web search feature so obviously this is very powerful for doing research finding the latest news connecting the AI to the internet because otherwise without that most of these models are very outdated they're only trained up to a certain amount of data so using the web search feature we can actually get information in real time so what I'm going to say today is what happened in today's AI headlines right and we'll give that prompt to quen 2. 5 Max then we're going to test the same thing inside deep seek let's start a new chat here we go and make sure we have search selected and then we're also going to do the same inside chat GPT however look at that it's blurred out we cannot use the search inside chat gbt now mind this is a pay tool I am paying for chat gp01 for the privilege of using this model but I can only use the search feature inside chat gbt 40 it is not available inside the chat here so deep seek has come back to us deep seek search VI sorry deep seek search service is busy please disable search or try again later right so that completely failed Chachi P1 doesn't even given the aution and then quen 2. 5 Max is absolutely smashed it look at this and we've got a bunch of ideas right here and you can see this is pretty up to date so for example daily update January the 30th 2025 which is today we've got the news headlines so quen 2. 5 Max didn't just work which by default means it wins but also it found us the latest headlines and therefore we are smashing it so you got to bear in mind here logic and reasoning model failed both tests fail both tests quen 2. 5 Max not a reasoning model not a logic model but absolutely smash both for these and is available for free next up what's what
3:31

Image Generation Showdown

we're going to be testing is deep seeks image generation versus Quin 2. 5S versus his chat g1's image generation so what we're going to do from here is we're going to click on image generation inside qu 2. 5 Max and we're going to put cat eating a banana all right and that's the same prompt that chat GPT is going to get so we'll put image of cat eating banana and we'll do the same inside this multimodal model from Deep seat which is posted on huging face because it is too powerful to host locally for me so we'll keep the other options default we'll hit generate image and that is generated now as you can see the black line is going up so quen 2. 5 Max has actually failed on us look at that error we'll give it another go but I don't think that's going to work at all let's see what we get back it is generating the image but it did fail the first test now chat gp01 cannot generate images so we're going to switch this to GPC 40 as well just to make this a fair test all right so we'll wait for that to load and then we have the image back from Janus and it did everything first time round so no messing around there and if we open this up that is a cat eating a banana I mean it's absolutely smashing that banana to be fair let's have a look through the other images he's having a great time there this one is getting his teeth right stuck in the only thing that I don't like about these images is that the images are very pixelated they you know like it's not quality is it it's not quality image and it's not very convincing either if we go over to quen 2. 5 Max that's how the better go it to be fair that does seem realistic I think that could fool a lot of people if you didn't know it was Ai and if we go inside chat gbt photo I don't like this at all I wouldn't use that on my content because it's just so cartoonish if I'm sure you can ask it to generate a different style so we could say like make it to graphic but you're going to have to do that every single time right so out of all of these if I had to put them in order I'm going to say quen 2. 5 Max smashed it Janus Pro 7B from deeps AI I've actually got another video on this if you want to check out has come in second and then even when we ask it to come up with something more like Photo style it's still not quite a rare it still doesn't feel realistic little Pro tip for you here actually if you really want to get good outputs for images I would recommend Gro it's pretty fast and it gives you four different options at the same time that's you have a drawback from quen 2. 5 Max which is that you just get one image same with Chad GPT whereas with Janis Pro 7B you get five different images to choose from even if you wouldn't choose any but you can see the images from grock here and aside from the banana floating in the air I would say the gro would win that any day of the week right and Gro is using flux as far as I'm aware for generating those AI images super
6:29

Video Generation Capabilities

impressive stuff so what we're going to do now is we're going to test inside the new chat we're going to go to video generation here and this is one I prepared earlier because it takes so long for Quin 2. 5 Max to generate videos so just bear in mind like it did take about 20 minutes or 30 minutes to generate this but we basically use the prompt cat eating biscos and here's what we got back multiple cats eating biscuits that's what a time to be alive you can automate all this with AI what a Time to be alive inside deep seeky can't generate videos so it's not really comparable but we can use Sora inside chat gbt if you're on the $200 per month plan now bear in mind quen 2. 5 Max is free deep SEC car1 is free if you want to generate images you have to be on the chat GPT Pro Plan and that costs $200 a month AK $2,400 per year that is a price I pay to entertain you folks at home now Sora is a pretty cool platform to be fair like you can see you get loads of different image videos you can see what people are creating in real time like that does beat Quin 2. 5 Max but if we go over to here we're just going to go for like a quick 5sec video we'll plug that in catty in biscos let's see what we get back and it's taking its sweet time so we might come back to that in a second the final thing that I want to compare between all these models is whether you can use them locally right so here's the thing that you should also know about and it's super important to get this right by the way we just got this video back so let me cover this first and then we'll come back to that so we got cat eating biscos honestly that's probably the most realistic video out of all of them I would say that wins by a long way and it's given us four different options so it is superior to quen 2. 5 M I can see why it costs a lot more money I probably wouldn't use this as Boll but it's impressive and bear in mind like we just generated that at 480p you can generate it in higher resolutions which means that the videos will look better so out of all of them open AI Sora winds then Quin 2. 5 Maxs and obviously DC cannot generate
8:40

Local Hosting and API Access

videos now the final thing that I want to come on to is whether you can host these locally okay so if you have a llama downloaded which is completely free to download you can just download it right there and then you type in deep seek r one this is a massive Advantage for DC go you can just go to your terminal you can run you can make sure you have Al llama 2 running and then if we open this up you've now got deep SE G one ready to go with the distilled model right so for example if I say what model are you we can just run deep SE car1 locally and you can see here it comes back straight to us now that is a massive Advantage let's be honest because you can run this offline you can run it locally you can access this whenever you want and also you know some people are concerned about privacy so this will probably be a better way to use this now for quen 2. 5 you can download but the latest model AKA Quin 2. 5 Max is not available right so you cannot host that locally with a llama and chat chbt is not available locally either right and so if you're looking for something local that it's pretty lightweight cuz this is quite lightweight running on my laptop you know I can do up stuff in the background have this just running when I need it to like it's really good for that is a massive advantage to deep SE car one the other thing as well if you go to visual studio code and I'm just showing you this as an example we can actually use the API from Deep SE G one completely for free right so if we go to rot code right here I've got a video on it from yesterday and then if we go to open router plug in your open rout API key right which you can get from open r. and then you can plug in your API Keys you can just grab a new one from this section right here and then when you plug that into Visual Studio code with rot code or client or whatever you use you can actually select deep se/ deep seek hyen R1 fre this is amazing because now you get free API that you can code and build projects with for example you can see here we built a snow day calculator website earlier today we built a to-do list app didn't cost us Penny because we're using the free API from Deep seek so there's all sorts of amazing stuff you can do with the free API when you pair it with these models like visual shudio code now if we go inside the model section here you can use chat GPT but it's using Jack GPT 40 latest which obviously is going to cost you money right it's not going to be for free and then also if we type in quen here we can get access to Quin 2. 5 there's different models here you can get a free version of it so you get a Quin 27b for free as well but you don't get Quin 2. 5 Max for free on the API right so if you want a free API deep C car1 is smashing it if you want to host it locally deep C car1 is flying right but for most things for the average Joe I'm going to say qu 2. 5 Max does everything you need to right pretty impressive model great for Content creation HTML can code games is quite fast to respond it can create images videos Etc pretty good stuff chat gp01 is super powerful if you're running a business right so for example like the video options you know chat GPT operator these are like nice to have if you're running a business and you know you're not too fasted about the prices a deep
11:56

Conclusion and Recommendations

see car1 great for running locally so again quen 2. 5 Max for the average person watching this probably the best model so thanks so much for watching if you want to get access to a community of 3,700 AOS and a free AI course along with 53 AO tools and 200 free chat Che PD proms feel free to get that link in the comments description and if you want to get a free one to1 SEO strategy session Link in the comments description we'll show you how we take websites from 0 to 145,000 V this month and generate hundreds of thousands of dollars in sales on autot on this free link building acceleration ient you get a free SEO domination plan the secet SEO link boing answer any questions you have one to one you learn the best link boing strategy for your website plus ranky competitive building and 10x SEO traffic based on what's working for us and our happy clients like you see right here feel free to boot that in appreciate you watching byby

Ещё от Julian Goldie SEO

Ctrl+V

Экстракт Знаний в Telegram

Транскрипты, идеи, методички — всё самое полезное из лучших YouTube-каналов.

Подписаться