China’s GLM 5.2 VS Claude Code: Who Wins?

GLM 5.2 destroys Claude Code. So today we're going to be testing them out side by side, seeing how they perform. This is actually built with GLM 5.2, as you can see right here. So you can build out some pretty amazing stuff. Here's another example of what we created. We even built out a full 3D game, as you can see right here. And there's all sorts of interesting things you can do with this that we're going to run through today. And I'll show you exactly how it works, plus how it performs versus Opus A. The interesting thing about this is if we have a look at the benchmarks of GLM 5.2, which just got released, so we'll run through everything today and we'll compare them side by side. And I'll show you what we've built and how they compare versus Kimi K217 VS 4.8. So this is the new announcement. And bear in mind as well that this is open source. So it's available on Hugging Face. That means people can run it for free. It's MIT licensed open source project. And you can also run it on the coding plan. The other difference between Claude and GLM 5.2 is that GLM 5.2, you can actually plug into Hermes or your other AI agents using the coding plan, which is pretty amazing in itself. You can't do that with Claude. You can't use that inside your agents without paying for the API, which obviously costs a lot of resources. And so you can see here it's hosted on Hugging Face. It's available. And also on Design Arena, this is pretty shocking. It has scored number one on Design Arena. So it is outperforming Claude Fable 5 on Design Arena, which is wild when you think about it. Claude Fable 5 was so powerful, people took it down. Right. It literally just got taken down this week. GLM 5.2 doesn't matter, mate. We've plugged it in. This is a Chinese model, by the way. It comes from Zed. Claude Opus 4.6 is number three on this list. And you can see how it goes down. So, you know, GPT 5.5 is not even in the race here. And everything that we've built out of this does look pretty nice. I'll come on to that in a second. And we'll show you some tests. Also already on Olama, which is great as well. So you can use it on Olama. And we can see how it holds up on the benchmarks right here. We can cover more on the benchmarks. I just want to show you what we've actually built with it because it's kind of fun to look at this stuff straight away. So we have Kimi K 2.7. We have GLM 5.2. And we have Opus 4.8. Which one built the best first person raycaster engine, which is kind of like a, you know, a textured maze essentially. So let's run through these and we'll see how they perform. So first of all, this was Kimi K 2.1.5. It's actually super smooth. Like when you go around here, look how smooth that is when you're using it. It's pretty nice, right? So this is another alternative that's open source and just dropped from China. Then let's have a look at this one. So this is the option from GLM 5.2. It's not bad, but at the same time, it's not great. So if you have a look, you can see that it kind of feels a little bit buggy when you move around. And it's very hard to control that as well, right? But it's exciting. I mean, it is cool. Now, if we have a look at this is Opus 4.8. And I would say this probably created the best one. It's still a little bit hard to control as we're running through it. But I would say that is probably the best option. So so far, I would say Opus 4.8 wins on the first version. And Kimi K 2.7 nailed it as well. Now what we actually did is we leveled up the Raycaster and we added some interesting stuff in here. So if we open this up, as you can see, this is the output from Kimi K 2.7. Kind of reminds me of like a, you know, a Mega Drive game back in the day. And you can see an example right there. Pretty hard to control, but it looks cool. Now, this is the one from GLM 5.2. Actually pretty good, right? Actually pretty good. I would say it nailed it, to be honest. So if we go through here, we can also see a map in the top right as well, which is super useful. And then if we turn around, we will see things popping up as well as we go along. So pretty cool game right there. A lot of fun to play as well. And that feels like the most realistic on C. And then we have Opus 4.7. I would say Opus 4.8. I would say that looks the nicest so far out of every one that I've seen. It's also the most playable game, to be honest with you. So so far, Opus 4.8 is still winning on those games that we're creating. Now, let's have a look here. We have GLM 5.2 working on flying through a neon city. So that is Kimi K 2.7. Doesn't seem to work at all. This is GLM 5.2. Definitely works and we can fly through the city here. It's a little bit buggy. It doesn't feel that nice to use, but that's an option. And then if we have a look, this is Opus 4.8. Opus 4.8 has nailed it again. So that's three out of three. And this actually surprised me because we gave it like different tests this time and it's actually working. Whereas, for example, when I was using that with, I did another test on this the other day and GLM 5.2 was clearly winning back then. So it's interesting to see, okay, if you run different tests, you get different results and you get different outputs. But so far, Opus 4.8 is winning on all of this. Now, this is like a racing game. As you can see, Super Buggy from Kimi K 2.7, Super Buggy from GLM 5.2. And this is Opus 4.8. Still quite buggy, but probably the most playable out of all. But yeah, not a big fan of any of those. And then we have the solar system that you can orbit. So if we have a look here, this is Opus 4.8. This is GLM 5.2. And this is Kimi K 2.7. Now, if you look at this, Kimi K 2.7 is really buggy. I would say that GLM 5.2 is probably the best. Cloud Opus looks the smoothest, but it's not labeled or anything like that. And it doesn't have any options. So if you look here, you can actually click on these buttons. You can zoom in, you can zoom out, you can drag it, etc. And I would say there's a lot more customization on the output from GLM 5.2. So that is the first one that GLM 5.2 has won. So far, Opus has won. I would say it's won two tests. I would say this one is a draw because none of them were good. This one, GLM 5.2 has won. Now, this is interesting as well. So what we actually created here is like a voxel runner game. So if we have a look here, this is Kimi K 2.7. Pretty nice. Not bad at all. If we have a look, this is GLM 5.2. I would say that is a much better version, right? A lot more fun to play, etc. And then we have Opus 4.8, which is like super basic. And so if you look at this, actually GLM 5.2 wins this round. So GLM 5.2 has scored and won two. Claude Opus 4.8 is won two. I would say Kimi K 2.7 hasn't won any of them so far. Now, if we have a look at this test, this is similar to the Solar System one before. This one is not that great from GLM 5.2. Kimi's not bad. But obviously, if you look at this, like Opus 4.8 has crushed it, right? It's nailed it. So that's three versus two for Opus 4.8. For this one, this is liquid in a bowl. And we did a test. And it should hover as we move the mouse. And if we have a look here, for sure, GLM 5.2 has won this one, right? It's the most visual, it's the most interesting. We can change the theme over here. It's a lot more fun to play with. If we have a look at Opus 4.8, super boring, not interesting at all. If we look at Kimi K 2.7, super boring, right? So three for GLM 5.2 and three for Opus 4.8. Now, let's have a look at the rest of the benchmarks here. So GLM 5.2 is winning on the benchmark for Design Arena. And also, there's two things to note here. So it is open source. Claude is not open source. And then additionally as well, you can plug GLM 5.2 into your agents. You can't do that with Opus. So in many ways, like GLM 5.2 is doing pretty well. Now, if we look at the agent arena, if you have a look here. So if we look at this list, obviously, Fable 5 is top. Opus 4.8 are top at the agent arena. But the thing is, right, that you can't use OAuth with your agents. So if you were using agentic capabilities with Fable 5 or 4.8, you would have to use the API, right? If you wanted to use this with, for example, Hermes or OpenClaw. Whereas, for example, with GPT 5.5, which is 6 and 7. Or, for example, with GLM 5.2, you can use OAuth and you can log in to your agents. So I would say on that particular test, Clawed is the most powerful for sure. And it's beaten Clawed by, I mean, it's beaten GLM 5.2 by a long way. And also one thing to note here, this is pretty interesting actually, is like you can use Clawed code with GLM 5.2. So you can plug in the brain of GLM 5.2 on cloud with Olama into Clawed code. The same for Codex, the same for Hermes agent, and the same directly inside the chat. Now, if we look at these benchmarks, let's have a look where we're up to. So GLM 5.2, 62.1, Opus 4.8, 69.2 on SW Bench Pro. Terminal Bench, GLM 5.2 is just under Opus 4.8, 69 versus 48. So if you look at all of these benchmarks, like GLM 5.2 is up there. It's definitely beating Gemini, but it's not quite the same level as Opus. And I think I've found that as well, is like Clawed still feels a lot smoother to me when it just comes to day-to-day tasks. But it's right up there. And bear in mind as well, like GLM 5.2 is cheaper on the coding plan as well. So these are some of the tests that we've done. We also did a website landing page test here. So if we pull these up, this is a version from Clawed 4.8. Pretty basic, pretty, you know, not that interesting. And GLM 5.2 created something a bit more animated and a bit more interesting. You can see the nice colors here, etc. So I would say it won on that test as well. And this one, it definitely won. So this is a sort of neon game here versus Clawed Opus' like super basic game. But yeah, so much cool stuff. So I would say, you know, if you're looking for the best of the best, you're still going to stick with Clawed Opus. But if you want something that's cheap, that's affordable, that you can plug into your agents, that's doing really well in design, then I would go with GLM 5.2. And I would say, you know, if you were choosing between Kimi K 2.7 and GLM 5.2, I think GLM 5.2 is an absolute no-brainer, to be honest with you. But yeah, this, I mean, look at all this cool stuff that you can build with it. It's pretty interesting what you can build and how you can build with it. It is fun. Now, we've actually already plugged it into our agent operating system as well. So if we go to GLM 5.2 over here, we've got the chat, and we can switch between different agents over here, and then inside the workspace as well. Everything that we build just gets saved inside here too, which is super useful. So we can come back to everything that we've built, play it later. But yeah, very impressive model. And bear in mind, this is open source. Whereas, for example, you look at stuff like Fable 5 and it's getting shut down, right? That's the difference. So if you want everything in one place, so we've built out GLM 5.2, we're plugged into our Hermes agents. We've got this amazing agent operating system where we control all our agents and plug them in in one place. You can get that all inside the AI Profit Boardroom. Link in the comments description or go to theaiiprofitboardroom.com. And inside here, basically, this is my AI automation community for helping you scale and learn with AI automation. And you might say, well, GLM 5.2 is not that good. But actually, you've seen the outputs. Like, it's definitely up there with Claude Opus 4.8. Is it better than Opus 4.8? No. But is it cheaper, more affordable? Can you plug it into your agents? Does it have advantages that Claude Opus 4.8 doesn't have? Absolutely. And also, if you're wondering, okay, like that Agent OS system, it looks complicated or, you know, I don't know how to set it up, et cetera. We've got the full zip file inside the agent, inside the AI Profit Boardroom. And you can see here loads of people setting up their own versions of this as members inside the community. We actually have over 182 pages of testimonials and wins like this of people just getting amazing results with this stuff. So if you want to join us, it's inside the AI Profit Boardroom. Go to the classroom if you want to get the system. And then from there, you can go to the agent operating system section, as you can see right here. We have a video tutorial. We update it daily with the new stuff. You can get a zip file to install it. And then we add new tutorials and new stuff based on what's actually useful for you. Also, inside the community, I personally answer these questions, which is great as well. So, for example, here, you can see John asked a question earlier today. We've already answered him. We already got back to him. I actually create video tutorials daily for the members here based on their questions. And then also, you can get access to all of my best trainings inside the classroom, including how to go from beginner to expert with AI automation. Inside the calendar, you can jump on weekly coaching calls. So we have four weekly coaching calls where you can ask questions, share your screen, get help and support in real time, et cetera. Inside the map, you can meet people in your local area who are using AI agents just like you. And this is all inside the AI Profit Boardroom. Link in the comments description or go to the AIprofitboardroom.com to get access. Thanks for watching.