Gemma 3 VS ChatGPT VS Grok VS Claude: Who Wins? 👀

Gemma 3 which is the new update from Google versus grock versus chap versus Claude which one is the best now if you're not familiar with what Gemma 3 is let me pull up an example right here there's actually a whole section inside the AI profit boardroom on exactly how it works Etc what it is but essentially this is the latest model from Google and Google said this is the most capable model you can run on a single GPU like you can see you can actually run it locally as well and you can see the benchmarks here so according to the chatbot arena is outperforming deep seek version 3 outperforming 03 mini llama 3 Mr Lodge and Gemma 2 and you can see it's right up there with deep SE car 1 so we're going to be testing it and comparing it versus Gro chat GPT and Claude And if you want to get free access to this is completely free to use then you can get access at aist studio. goole. com so let's get straight into this first thing I'm going to say here is build and you know what I'm going to take some prompts I previously tested just to give this a fair run so first thing that we can do is

we're going to do a little reasoning challenge all right I'm going to say there's a tree on the other side of a river in Winter how can I pick an apple all right and we'll test it for reasoning first of all so the main thing we're looking for here is can it think for itself I'm not going to use a reasoning model inside any of these other models I so we're just going to be comparing like the basic models simply because as far as I'm aware Gemma 3 is not a reasoning model but you can see how it's coming up with an answer here actually it's pretty impressive the answer has come up with like already you can see the reasoning step by step and it's given us multiple different solutions right so that's the main thing that we're looking for is can it understand and can of look at it from a logical way and then also can it problem solve for itself that's a really detailed answer and it's the simplest answer Apple don't stay on trees forever wait until spring summer or fall when the river might freeze or the water level is lower or you can simply wait across the boat right now let's have a look here has it actually figured out that apple trees don't have there we go right it says it's winter apples aren't typically rape or even on the tree in Winter the riddle implies there is an apple but it might be a trick all right so it's figured that out plus it's giving us loads of different solutions here pretty practical and also it's straight to the point so pretty good answer right there let's see what we got back from Gro is not bad it's not giv us as many ideas as you can see but still not bad at all chat says use a drone but it hasn't figured out that actually in Winter apples typically don't grow on trees so that's coming in last so far and let's just check out Claude and Claude figures out straightway say like it's winter and apple trees typically don't bear fruit in winter so when I look at this I actually think the most practical Solutions came from GMA 3 Gro did pretty well too Claude figured out straight away so that's coming in second so far we got GMA 3 coming in first Claude second grock I would say joint second chbt 40 didn't get it at all right now bear in mind this is a free model so you get free access to Gemma 3 as well which super powerful and we'll move on to the next example and we're going to say okay let

me pull this up I'm going to give it this prompt which I usually use for Content writing Link in the comments description to the Ari profit boardroom if you want to get all my best AI SEO automations we going to plug this into the chat here and we say create an SEO optimized article for this keyword equals and we put SEO training London we'll plug that in we'll do that inside Claude inside grock I'll probably have to sign into grock on there directly we'll continue with X and then we'll also do the same thing in the background with chat GPT I'm going to start a new conversation for each just so that it's got more context Etc all right we're logged into Gro as well let's plug that in so same prompt in each and basically what we're testing for is which is the best of creating content now usually my personal preference is using Claude seems to be the best when it's creating content but let's see what we get back so Claude SEO training London The Ultimate Guide so pretty nicely written right there let's grab the text see what word count we got out so it's come back to us with 1,000 words I think we did ask for how many words did we ask for maybe we didn't give it a word count all right so that's 1,000 words let's see what we got back from Gemma is still writing the content it's a bit slower the content is prettyy nice nicely formatted as well so for example why London needs specific SE expertise competitive really competitive Europe up against established Brands pretty nicely humanized let's have a look at Gro let's check this out this is super long article let's see how long this is that's coming out of 1,200 words Gro is Gemma 3 is still writing the only thing I would say with Gemma is it still feels like really a I style text right so for example if you read this first line are you tired of your website being invisible on Google do you feel like everyone is getting clicks while you're stuck on page two or Worse doesn't feel that humanized whereas if you compare it to Claude which is you're looking for the best SEO trainer in London that actually delivers results I've been where you are right if you're tired of the same recycled advice that doesn't move the needle let me share what actually works that is 10 times better than the intro from Gemma which still feels like AI fluffy this is not bad from chat gbt let's see how long the content is only 650 words so I saying the way this is written it feels a lot nicer than Gemma 3 but it's still not as good as Claude and then if we have a look at Gro let's see where we're at and you can see how it's optimized for the keyword as well so for example SEO training in London is front loaded in the title and also in the first line feels very humanized and it reads really nicely probably not quite as nicely as Claude but it's up there some I'm going to rate this is Claude coming in first Gro coming in second chat GPT coming in third and then Gemma 3 it really does feel like AI fluff if we read through most of that content so I want to stick with Claude for that in terms of being the most humanized and so far Gemma 3 is not doing that well but hopefully it gets better for it so let's keep going now and we're going to move on to the next prompt which is actually what I want to

do here is I'm to see how well humanized the content is right so usually I use zero GPT to figure this out and what we can do is plug in the content from each Ai and see which one comes out the most humanized so let's grab the content from Claude so we'll close that out put set text and we'll see how these perform all right so you can see here claud's content Really humanized Right comes out at 0. 9% AI detectable which is pretty impressive to be fair bear in mind it created a thousand-word article and it's not AI detectable at all let's check now how Gro performs so we'll do the same thing so many popups on Z GPT is crazy so we've got the content back and that is coming out at 15% so it's currently in second place that's the content from Gro let's take the content from chat chpt and then after that we'll check Gemma 3 so that's 19% all right so far in order Claude comes out 0% human AI detectable Gro comes out at 15% chat gbti came out at 19. 39% and finally let's check Gemma so it's actually come out in the order that I rated it as well so I did say before Claude came in first Gro came in second Chachi came in third which is pretty interesting so I think to the human eye as well once you create a lot of content with AI you just get a feel for whether it's AI written or humanized at all and there like I said Gemini felt sorry Gemma felt really AI when I was reading through it felt like there was a lot of AI fluff in there and zero GPT can feel that too right so it says your text is likely human written but may include Parts generated by Ai and that comes out at 39. 3 n% AI detectable so in terms of not just the humanization and the writing of the content but also the actual detectability of it you can clearly see the chat the Claude is the best for human writing humaniz writing Gro comes out second chat gbt comes in third and Gemma is really laggy behind by quite a long way right there now it might perform better when you're running it locally but I'm just testing it inside AI studio so you know you get the feel for it and as well it took two minutes to generate this content whereas I think Claude was a lot faster when it was generating that content plus it created more content for us as an output let's keep going now what else can we do so I'm going to now test it for coding

so I'm going to say create a self playing snake game using python with a simple GUI and I'm actually going to change that to HTML so we'll test that on each let's go back to Gemma we'll start a new chat here same for Claude gr the only thing I find with Gro is this ly average compared to Claude right I really only use Claude these days chat chpt is useful too but honestly these days like Claude is so powerful and so good at writing and coding and giving you outputs back quickly plus it's very reliable that I tend to use Claude more than anything these days but let's see what we get back from Gemma now Gro has already come back to us so I think we should be able to run that yeah so we can preview that so we got a self playing snake game there that is working perfectly so good job from Gro Let's test out Claude is still coding so is Gemma 3 so let's check out chat gbt now we're going to click on preview over here we'll see what this how this performs as you can see if you compare the output of grock versus the output of chaty see how this is a lot slower and a lot more laggy compared to Gro which is like way faster when it's doing these outputs so there's still a big difference here between these models so far Gro is winning Let's test now Claude is quite interesting because it actually thinks more about the UI so you see how here the UI is nicer design it's got more color ranges we can actually change the speed so we can dial up the speed or dial it down we can also pause it reset it and control it so far it's good but let's see if it actually works and that is working perfectly all right so far Claude is winning on that respect but let's test out now we've got Gemma can we preview it directly inside the chat here I don't think we can more options Branch copy text Mark down no we don't seem to be able to run it so what I'm going to do from here is I will grab the HTML we'll copy that to clipboard and we'll go over to live weave which is a really useful tool for plugging in HTML we'll plug that in so to be fair that is pretty good but it doesn't work look at that game over let's see if it works again now or if it's just going to get stuck every time does it work no look at that it's not playing itself it's just running into a wall over and over again all right so if we compare these models GMA is coming in last because it doesn't actually work like you can see Claude is coming in first like you've got the UI you've got the controls you can change the speed it's even got a title and the colors plus the design is much nicer the front end I would say it's coming in second just because it's a bit faster than chat gpt's output and chat GPT comes in third so Gemma is not really performing that when it comes to for example coding content creation content humanization it does perform better on reasoning versus these other models but that's about it from what I've seen so far so if you're thinking okay should I switch to one to gemmer instead of these other tools I would stick with Claude the other thing to know is yes Gemma is free but so is Claude right the only thing that you

might use Gemma for that is an advantage versus everything else is some when you can run it locally right so for example with Claude you can't run that locally Gro you can't chat gbt obviously you can't run that locally either so Gemma is winning there right if you want to set up Gemma 3 running it locally then you can go to our llama go to Gemma 3 like you can see and then installed the local model yesterday so I'm just going to open up a llama here right so you need to make sure that you have a Lama 3 running in the background so you can see the icon for llama right there a llama and then if we go over to terminal here and we type in and we just need to make sure we have our llama 3 running all right and then if we go over terminal there we go you see how that's running now and we set okay what model are you we'll zoom in and you can see it says Gemma right so if you want to run this locally it offline Etc then Gemma is the only option out of all these tools so it comes in first there and then also if you want a free API Claude doesn't have a free API neither does Gro neither does chat chpc I've actually got another video on AI agents and building them with Gemma 3 and that's one of his biggest advantages as well so the way that I usually set up Gemma 3 from what I've tested so far is I will go on to open router right and I'm going to use open router for the API and then use that to build stuff out so for example if I'm plugging this into na1 let's pull this up so if you want to build like one click AI agents we can go into na10 and then when we're building out a workflow for example let's pull this one up all right so for example like this one then we can use open rout and Gemma 3 there right we have a bunch of templates for AI agenes inside the AI profit border so if you go to the classroom then you go to AI agents we have tons of different resources here so you can download all of these and import them into na10 and then we also have a bunch of others as well right now if you want to use this inside na1 let me show you exactly how easy this is so we'll go to create workflow and for here we're going to put in chat trigger after that we'll type in AI agent like and then inside the chat model here we're going to type in open router from open rter we'll select Gemma so we'll go with Gemma like so you might have to switch routinely so just be careful which one you select sometimes you know when it's a new model it doesn't seem to work perfectly and then from here we can say okay are you working and again it says here we go let me open this up here so it says yes I am always working as a large language model I don't clock in or clock out right so it's actually got quite a bit of Personality as well when it repes so if you compare Claude versus Gro versus chat GPT versus Gemma 3 then Gemma 3 is pretty good for building out free AI agents you can get a free API inside open router and then you can plug that into na10 whereas the rest of the apis you're going to have to pay for so that's one of the biggest advantages so

that's all I would use GMA 3 for local and free apis but I wouldn't use it for example like reasoning or coding or creating content or humanizing content directly inside the chat so thanks so much for watching that is the end of this video If you want to get access to all of the agents workflows templates saps video notes Etc from today feel free to get that inside the AI profit boardroom this is basically a community designed to help you save hundreds of hours to make more money with AI and inside here you'll get access to our best email and content automations social media automations AI agents web automations scraping AI SEO automations Q& A call recordings saps and a crash course to save you hundreds of hours top of that if you have any questions feel free to ask inside the community here so you can ask any questions you want inside the community if you get stuck and if you want Live help on a weekly call just jump on the coaching calls right here as you can see feel free to ask us any questions you have Link in the comments and description to that and also if you want to get a free suf strategy Session One to1 feel free to put in we'll show you how we take websites from 0 to 145,000 vids this month and generate hundreds of thousands of dollars in sales and multip on this free link building acceleration session you'll get free SEO domination plans you get a custom tailored SEO link building plan so you can generate more lead sales and profits from your website you'll discover the secrets as your link building answer any questions you have you learn the best link button strategy for your website plus how to our rank your competitors based on what's working for us feel free to get that link in the comments description appreciate you watching