The script reviews the new Grok Build coding agent and CLI, saying it’s a major upgrade over the old version and the best results the creator has seen from the same prompt, especially for games and 3D builds. It showcases multiple outputs (a dark first-person game, clean websites, 3D/MMO/RPG-style demos, animations like glowing fish trails and lava-lamp liquid that follows the mouse) while noting quality is inconsistent and sometimes disconnected from prompts. The creator compares Grok Build against Kimi Code, GLM 5.2, and Claude Opus, claiming Grok Build’s game beat Claude’s for the same prompt and outperformed Kimi and GLM on some tests, though GLM looked better for certain animations. They highlight OAuth login and that X Premium Plus/Super Grok subscribers get access, plus parallel sub-agents, self-check loops, and integration into an agent operating system alongside other models, then promote an Agent OS package via AI Profit Bot.
00:00 Grok Build Surprise
00:37 Touring Early Demos
01:12 OAuth and CLI Access
02:08 Studio Media Features
02:36 Beta Announcement Context
03:11 Side by Side Comparisons
04:37 Animations Mixed Results
05:50 Parallel Sub Agents
06:22 Dragon Realm Benchmark
07:02 Self Check Improvement Loop
07:58 Using Grok in Agent OS
08:48 Agent OS Offer Breakdown
09:43 Community Coaching and Close
Full transcript
Grog build is actually beast mode and honestly everything else I've previously tried with grog build was pretty terrible but this bad boy is looking nice. I'll compare it against what we've created with other models but this is literally the best build I've seen out of everything I've tried to create with this same prompt. Now if we have a look as well we've got another game right here let's check this out. So you can see it's kind of like a very dark first person game.
I'm not really sure what it's trying to do and again we'll compare that against some other stuff that we've built in a second. Let's take a look at some other builds here we've got a website that's looking pretty nice and clean that's looking decent. We have this really cool 3d it seems pretty cool building like 3d games. That's what I've seen with this like websites you know similar to everything else I've seen but coding out for example games and 3d stuff really really cool rpgs etc nice and this is actually quite fun to play it's hard to switch off.
So you can see some of the cool stuff that we're building with this and this is all with grog build now the other thing that shocks me about all of this is that you can just log in with grog build on or off so if you're already subscribed to twitter no problem mate you actually have the cli included with your twitter subscription which blows me away. So we've been building this directly into the agent os earlier today one thing I will say is like I was testing out the old grog build versus the new one and this is a huge step up. Now also what's really good about this is there's a cli so you can actually build this into your agent operating system as a cli and have it sit in the next alongside for example claude or hermes or gemini whatever you want which is pretty nice as well and not only that my friends you can also plug in grog with your twitter subscription again directly into hermes agent and then you can use it inside the studio so not only can it code out using grog build in the cli not only can you build the cli and add it alongside all your other agents inside your operating system but also on top of that you can generate voice notes you can generate videos you can actually train grog to create videos for you so we've got it trained to create avatar videos like you can see and then also we have images that it can generate too and it's a pretty good alternative to minimax on hermes agent. Now elon actually posted earlier today that he was looking for feedback on this sort of stuff honestly from what I've seen it is good it is good so they've said and this was announced on 25th of may like today we're launching an early beta of grog build a powerful new coding agent and cli for professional software engineering and complex coding work available now to all super grog and x premium plus subscribers now I will say some of this stuff not that great so I want to be 100% honest for you so if we have a look at this one for example I mean what even is this like it's so dark I can barely even use it let me compare that side by side versus kimi code so kimi code actually built something similar for us which you can find over here and if we open this up it's a lot nicer a lot better right not not perfect by any means but it's getting there and it's a lot nicer than what we had previously so for some stuff it's really like mediocre but for other stuff it's pretty amazing let's take a look at what else he built also this is important to note so this is the game that we created with grog build as you can see right pretty cool looks nice nice and clean fun little design etc pretty fun to play now if we compare that versus what we created with opus 4.8 it actually created something better which I've never seen with grog before like in every single test I've ever done claude wins but this time you can see how boring claude's build out for the same prompt is compared to what we created with grog build over here right and that is very interesting this is the first time I've ever seen grog get anywhere near let alone beat claude opus on tests I would genuinely say it probably beat all of these right probably I think it did outperform kimi k217 and glm 5.2 on the same test which is impressive in itself and again you might say well I don't want to use api blah blah blah but you can use your twitter subscription for this now this was an animation of hundreds of glowing fish with flowing trails pretty interesting now I also built some stuff that I thought well I'm not sure about that so I wouldn't like switch everything over like for example this orbit diagram is okay so it's like a inner cellular system now this one is pretty bad and this one is like kind of meh but not that interesting same for these they're kind of like interesting but weird abstract creations whereas if you have a look at what for example glm 5.2 created using the same prompt it's a lot nicer and it flows a lot more so in some strange ways it just feels kind of disconnected from the prompt when using grog on certain things I also tested it for a lava lamp animation as you can see that was pretty cool and again like we tested this on the new version of grok build versus the old one the old one was was terrible like wouldn't waste your time on it but this one is much nicer so it's definitely stepping up here's another version and it's pretty cool so this liquid follows the mouse and then you can change the colors over here and reset the ball as well now also what you get with grok cli is for larger tasks grok build will delegate work to specialized sub-agents that run in parallel so basically you can have a team of ai agents working on one task at once so that's how we've managed I've only tested this for like literally an hour and I just kind of walked off and gave claude access and then came back later but from what I can see you can give it one sentence then you have a full factory of eight sub-agents working together and this is how we've managed to build so many things so quickly across so many different projects now if we have a look at this example the dragon realm so it looks super nice it's really smooth to use very very quick to navigate and it works basically perfect right now the graphics are not great if you look up in the sky that's supposed to be a dragon so just want to be honest with you there but if we compare that versus what we got on the same test with kimmyk217 you can see how basic this looks and again it's not that smooth in fact if we try to walk forwards we go backwards and it's a lot darker the ui is not as nice it's just not as smooth as a version that grok build created now also with those sub-agents it's got it can also check its own work so how this works is basically a loop where you plan it the sub-agents start creating it it self-checks itself ships it and then uses that self-check to improve it on the next loop so the more you use it the better it should get in theory you might also say well you know grok it can't beat anything or it can't build anything but so far it created some pretty nice stuff which i showed you you also might say well i need to be a coder to use cli or to use grok build i wouldn't let that put you off it's just like using claude you talk to it inside the cli it gets stuff done for you and you just talk to it and message it like a friend so it's pretty chill other people say well i'll wait until these coding agents settle down or i'll stick to one thing they're not going to settle down anytime i think it's fun to try this stuff and test it out and see side by side grok build genuinely surprised me for a lot of stuff it created now also the cool thing about this is if you have an agent operating system like us where we've got claude and hermes and gemini and everything together we can now select grok build we're not using the api we're just using the subscription they already had so it doesn't cost anything extra and then inside the workspace we can just preview everything that we've built so it will get saved along with the history of what we've created and we can come back to it anytime which is pretty awesome so that's a great way to use grok build you know you can plug it into your hermes agent so you could have it inside a cli here and also it's going to plug into your memory system as well which means it's instantly got context even though you're using it for maybe the first time today or some of you will have already been using it but if you're using it for the first time like i am then all of a sudden it's got as much context as the agents that i've been working with for months and that's what we like so if you want the whole setup that i've used you know where it turns grok build claude glm kimmy and the rest into one system with shared memory and one dashboard you control every new model that drops makes a whole factory more powerful automatically and you get the full agent operating system zip file with every prompt and every model you get the obsidian memory setup so your agents know your business yet coaching calls where we can set up together step by step and a 30-day roadmap daily tutorials a member map and 24-7 community plus an awesome community of people who are actually implementing this and getting it done which we love you see how many people are building their own agent operating systems and mission control systems here so that's what we like so if you want to get that it's all inside the air profitable link in the comments description or go to the air profitable.com and inside the classroom if you go to the agent operating system section you can get that system we update it daily we give you the new updates you get the zip file you get a video tutorial and also when new tutorials come out and new cool stuff comes out we share it inside the community like you can see before step-by-step guide and video tutorial you can also ask questions inside the community and we help you i personally answer those questions too inside the calendar four weekly coaching calls where you jump on call share your screen ask about your setup if you're complete beginner you can go from beginner to expert over here inside this beginner to expert course and you get the new updates and agent os over here inside the map you can also connect with people locally near you and we've got people all over as you can see right here so hope to see you inside there cheers for watching bye
More episodes