Elon Musk's Grok3 Just STUNNED The Entire AI Industry (Beats Everything)

So it's quite clear that when Elon Musk said that Grok3 is the world's smartest AI, he wasn't actually jumping on the hype train. Today he unveiled Grok3, which is by far the world's smartest AI across a range of several different benchmarks. In today's video, I'll get through all of the announcements that you need to understand for Grok3 and show you guys why this is actually the world's smartest AI in its current form. So one of the first things most people of course going to want to look at is of course the benchmarks. And if we take a look at Grok3, the non reasoning model, we can actually see that those benchmarks are pretty incredible across the board. We can clearly see that Grok 3 and Grok 3 mini actually surpass recent state of the art models like Gemini 2, Deepseek V3, Claw 3.5, Sonnet and the recently updated GPT4O. This is something that is truly, truly incredible. And even if you guys think the benchmarks aren't important, later on the team actually goes to show why. They looked at new benchmarks, tested Grok 3 on those, and it still manages to excel. So it seems like all of the training, like this huge training run that they managed to do has actually managed to make the model that much smarter. And these scaling laws are still performing very well.

You know, the model is still currently training actually. So this is a little preview of our benchmark numbers. So we evaluated Grok3 on three different categories. General mathematical reasonings, on general knowledge about STEM and science, and then also on computer science coding. So Amy, American Invitational Math Examination hosts it once a year. And if we evaluate the model performance, we can see that the Grox 3 across the board is in a league of its own. Even his little brother, Grok 3 mini is reaching the frontier across all the other competitors. So you will say, well at this point, all these benchmarks, you're just evaluating the memorization of the textbooks, memorization of the GitHub, repos. How about a real time usefulness? How about we actually use those models in our product?

Other thing that I do love about the Grok3 team is that they also managed to put their model into the Chatbot arena. If you aren't familiar with the Chatbot arena, it's not quite like a standardized test, essentially a test where you have two AI models that give you a response anytime you ask an AI question. And out of the two responses that you get, you essentially pick which one you think is better. And it's a blind test, so you don't know which model is called which, you just receive a response, you click which one you think is better. And over time they showcase which model has been receiving the most number of wins. And currently the number one model on the chatbot arena is actually Grok3. So even in blind tests where people aren't biased by the names of the model, Grok 3 is something that is clearly winning. And remember, this is just the non reasoning model.

We actually kicked off a blind test of our Grok 3 model, code named Chocolate. It's pretty hot. Yeah, hot Chocolate. And been running on this platform called Chabo arena for two weeks. I think the entire X platform at some point speculated that this might be the next generation of AI coming away. So how this Chatbot arena works is that it strip away the entire product surface. Right. It's just raw comparison of the engine of those AGIs, the language models themselves, and place the interface where the user will submit one single query and you get to show two responses. You don't know which model they come from and in the end you make the vote. So in this blind test, Groq 3, an early version of Groq 3 already reached like 1400. No other models has reached an Elo score head to head comparison to all the other models at this score. And it's not just one single category, it's 1400 aggregated across all the categories in chatbot capabilities, instruction following, coding. So it's number one across the board in this blind test and it's still.

Climbing, so we actually have to keep updating it. So it's 1400, about 1400 and climbing.

In fact, we have a version of the model that we think is already much better than the one that we tested here. We'll see how far it gets. But that's the one that we're working on or talking about today.

Yeah, so actually one thing, if you're using Grok 3, I think you may notice improvements almost every day because we're continuously improving the model. So literally even within 24 hours you'll see improvements.

So now of course is the juicy part, the reasoning models. So these models are basically models that think for a extended period of time rather than just giving you an instantaneous response. If you aren't familiar with why these models do this, it's because it allows the models to think for a longer amount of time, giving them access to better and higher quality responses, enabling them to be more accurate and tackle more Complex problems. This is something that we've moved to do as an industry because this is something that you know, is providing promising results and will likely lead us to truly smart AI. Now, when we look at Croc3's reasoning capabilities in terms of the thinking models, those two also managed to surpass even the recently debuted O3 mini that many people heralded as the smartest AI on the planet. But now that is unfortunately number two.

Yeah, okay, so let's see how GROK does on those interesting challenging benchmarks. So yeah, so reasoning again refers to those models that actually thinks for quite a long time before it tries to solve a problem. So in this case around a month ago, the graph 3 pre training finishes. So after that we work very hard to put the reasoning capability into the current Graph 3 model. But again, this is very early days. So the model is still currently in training. So right now what we are going to show to people is this beta version of the Graph 3 reasoning model. Alongside we also are training a mini version of the reasoning model. So essentially on this plot you can see the graph 3 reasoning beta and then graph 3 mini reasoning. The graph 3 mini reasoning is actually a model that we train for much longer time and you can see that sometimes it actually perform slightly better compared to the Graal 3 reasoning. This also just means that there's a huge potential for the graph 3 reasoning because it's trained for much less time. All right, so let's actually look at how it does on those three benchmarks. So Jimmy also introduced already. So essentially we're looking at three different areas. Mathematics, science and coding. And for math we're picking this high school competition math problem. For science we actually pick those PhD levels science questions. And for coding it's also actually pretty challenging. It's competitive coding and also some LEET code, which is some code interview problems that people usually get when they interview for companies. So on those benchmarks you can see that the gra3 actually performed quite well across the board compared to other competitors. Yeah, so it's pretty promising. These models are very smart.

So Tony, what are those shaded bars?

Yeah, so okay, so I'm glad you asked this question. So for those models because it can reason, it can think. You can also ask them to even think longer. You can spend more what we call test and compute, which means you can spend more time to reason to think about a problem before you spit out the answer. So in this case the shaded bar here means that we just asked the model to spend more time. You can solve the Same problem many, many times before it tries to conclude what is the right solution. And once you give this compute or this kind of budget to the model, it turns out the model can even perform better. So this is essentially the shaded bar in those box.

So I think this is really exciting because now instead of just doing one chain of thoughts with AI, why not do multiple all at once?

Yes.

So that's a very powerful technique that allows to continue scale the model capabilities after training.

So they also wanted to see if this was just a situation where because they trained it for so long and on so much data, was this something that was just, you know, overfitting and it's basically just memorizing parts of the test. They decided to test it on the newer Amy 2025 and the results were pretty surprising.

And you know, people often ask, are we actually just overfitting to the benchmarks? So how about generalization?

So, yes, I think, yeah, this is definitely a question that we are asking ourselves whether we are overfitting to those current benchmarks. Luckily, we have a real test. So about five days ago, Amy 2025 just finished. This is where high school students compete in this particular benchmark. So we got this very fresh new competition and then we asked our two models to compete on the same benchmark at the same exam. It turns out very interestingly, the Graph three reasoning, the big one, actually does better on this particular new fresh exam. This also means that the generalization capability of the big model is stronger, much stronger compared to the smaller model if you compare to the last year's exam. Actually, this is the opposite. The smaller model kind of learns the previous exams better. So, yeah, so this actually shows some kind of true generalization from the model.

So 17 months ago, our Grok Zero and Grok One barely solves any high school problems.

That's right.

And now we have a kid that just already graduated. The GROQ is ready to go to college. Is that right?

Yeah.

I mean, it won't be long before simply perfect. The human exams won't be hard. They'll be too easy.

Yeah, like, and internally, we actually, as a grok. Exactly. Continually evolves. We're going to talk about what we're excited about, but very soon there will be no more benchmarks left.

Yeah.

Now let's actually take a look at these reasoning capabilities in action to show you guys what these models can actually do.

Yeah. So like Jimmy said, we've added advanced reasoning capabilities to groq and we've been testing them pretty heavily over the last few Weeks. I wanted to give you a little bit of a taste of what it looks like when GROK is solving hard reasoning problems. So we've had two little problems for you. One comes from physics and one is actually a game that GROK is going to write for us. So when it comes to the physics problem, what we want GROK to do is to plot a viable trajectory, to do a transfer from Earth to Mars, and then at a later point in time, a transfer back from Mars to Earth. And that requires some physics that GROK will have to understand. So we're going to challenge grok, come up with a viable trajectory, calculate it, and then plot it for us so we can see it. And, yeah, this is totally unscripted, by the way.

That's the entirety of the prompt which was to be clarified. There's nothing more than that.

Yeah, exactly. This is the GROK interface and we've typed in this text that you can see here. Generate code for an animated 3D plots of a launch from Earth, landing on Mars and then back to Earth at the next launch window. And we've now kicked off the query. And you can see GROK is thinking. So part of grok's advanced reasoning capabilities are these thinking traces that you can see here. You can even go inside and actually read what GROK is thinking as it's going through the problem, as it's trying to solve it.

Yeah, we should. Like, we are doing some obscuration of the thinking so that our model doesn't get totally copied instantly. So there's more to the thinking than is displayed in.

All right, so this was the little physics problem we had. We've collapsed the faults here, so they're hidden. And then we see grok's answer below that. So it explains it. Wrote a Python script here using matplotlib, then gives us all of the code. So let's take a quick look at the code. You know, seems like it's doing reasonable things here. Not totally off the mark. Solve Kepler, says here. So maybe it's solving Kepler's laws, Kepler's law, numerically. Yeah. There's really only one way to find out if this thing is working. I'd say let's give it a try. Let's run the code. All right. And we can see, yeah, GOG is animating two different planets, Earth and Mars, here. And then the green ball is the vehicle that's transiting the spacecraft that's transitioning between Earth and Mars. And you could see the journey from Earth to Mars. And looks like. Yeah, Indeed, the astronauts returned safely at the right moment in time. So obviously this was just generated on the spot. So we can't tell you if that was actually a correct solution. So we're going to take a closer look now. Maybe we're going to call some colleagues from SpaceX, ask them if this is legit.

It's pretty close. There's a lot of complexities in the actual orbits that have to be taken into account, but this is pretty close to what it looks like.

Awesome.

Now Grok3 also entered its agentic era, which is something that I'm not surprised by. AI agents are essentially the theme for 2025 and beyond. And essentially they released something once again, I'm honestly surprised that all of these companies have name the product the exact same thing. They've called it once again Deep Research, or in this case they've called it Deep Search.

So today we're actually introducing a new product called Deep Search that is the first generation of algorithmic agents that not just helping the engineers and researchers and scientists to do coding, but actually help everyone to answer questions that you have day to day. It's kind of like a next generation search engine that really help you to understand the universe. So you can start asking questions like for example, hey, when is the next starship launch day? For example. So let's try that. If we hit the answer on the left hand side, we see a high level progress bar. Essentially the model knowledge is going to do one single search like the current racism, but actually thought very deeply about hey, what's the user intent here and what are the facts I should consider at the same time and how many different websites actually go and read their content. So this can really save hundreds of hours of everyone's Google time if you want to really look into certain topics. And then on the right hand side you can see the bullet summaries of how the current model is doing, what website is browsing, what source is verifying, and oftentimes actually cross validate different sources out there to make sure the answer is actually correct before it's output final answer. And we can at the same time fire up a few more queries. How about, you know you're a gamer, right?

Sure, yeah.

So how about what are some of the best builds and most popular builds in Password? Excel. Hardcore, right? Hardcore league.

You can technically just look at the hardcore ladder. Might be a fast way to figure it out.

Yeah, we'll see what the model does and then we can also do something more fun. For example, how about make a prediction about Martian madness out there.

Yeah. So this is kind of a fun one where Warren Buffett has a billion dollar bet. If you can exactly match the, I think the sort of the entire winning tree of March Madness, you can win a billion dollars from Warren Buffett. So like it would be pretty cool if AI could help you win a billion dollars from Buffett. That seems like a pretty good investment.

Let's go now.

Another neat feature of this deep search is that you can actually look at the chain of thought of the model. So if the model doesn't respond with something that you wanted, you can actually look at how the model reasoned through its search data to see how it came to that concl. I think this one's really useful because they talk about having the model be as transparent as possible and this is actually going to make it even more useful. So if you don't get the response that you do want, you can actually reason, look into the model sorts and figure out why.

And then in this case you can actually scroll through, actually reading through the mind of Grok. What informations does the model actually think about are trustworthy, what are not? How does it actually cross validate different information sources? So that makes the entire search experience and information retrieval process a lot more transparent to our users.

And this is much more powerful than any search engine out there. You can literally just tell it only use sources from X. It will try to respect that. And so it's much more steerable, much more intelligent than it really should save.

You a lot of time. So something that might take you half an hour or an hour of researching on the web or searching social media, you can just ask it to go do that and come back and 10 minutes later it's done an hour's worth of work for you. That's really what it comes down to.

So now if you are wondering how this AI is going to be rolled out, they actually speak about a new website called grok.com currently as of recording this video, the website is unfortunately down. I'm pretty sure the hype just simply broke the website. Maybe they didn't expect that many viewers, but it's basically going to be on grok.com, where they also have the super Grok, which is basically going to be where you can access the app dedicated.

On their website, one of the most advanced capabilities and earliest access to new features. So feel free to check that out as well.

This is for the dedicated GROK app and for the website.

Exactly. So our new website is called grok.com yeah. And you'll also find you never guess. Yeah, you'll never guess. And you can also find our Grok app in the iOS app store. And that gives you even more polished experience. Experience that's totally Grok focused. If you want to have grok easily available, one tap away.

Yeah.

And the version on grok.com on a web browser is going to be the latest and most advanced version because obviously it takes us a while to get something into an app and then get it approved by the App Store. And if something's on a phone format, there's limitations of what you can do. So the most powerful version of Grok and the latest version will be the. The web version@grok.com.

Yeah. So watch out for the name Grokfree in the app. Yeah, exactly. That's the giveaway that you have Grok free. And if it says Grok true, then Grokfree hasn't quite arrived for yet. But we're working hard to roll this out today.