SEO Gold Daily
← All episodes
Episode 15 · June 15, 2026 · 11:22

GLM 5.2: New AI SEO Agent is Insane!



GLM 5.2 + Hermes Agents: Build Ranking Websites & Videos in 30 Minutes (No Coding)

The script explains why GLM 5.2, released by Zed AI on June 13, is powerful for AI SEO due to its million-token context window and suitability for agentic tasks, and shows it being plugged into Hermes Agent to create teams of AI workers. The speaker demonstrates building a fast-loading, professional website with multiple pages and blog posts in about 30 minutes using agents like a researcher, writer, SEO optimizer, and a “judge” that iterates until quality improves, aiming to rank on Google and generate leads without ads. It also describes producing an AI-avatar video end-to-end (script, voiceover, look, editing) that can rank in search results, compares GLM’s flat $18/month coding plan approach to using Claude via API, highlights swapping models as new ones release, and promotes a free SEO course, strategy sessions, and the AI Profit Builder system and community.

00:00 GLM 5.2 Demo Promise
00:10 Model Drop Overview
00:29 Plug Into Hermes Agents
01:14 Website Built To Rank
01:36 No Coding Needed
02:38 Meet The Agent Crew
03:33 Website Walkthrough
04:36 SEO Proof And Case Studies
05:30 Agent OS And Video SEO
06:43 Claude Comparison Pricing
07:18 Swap Models Stay Ahead
08:31 Kanban Auto Assignment
09:14 Open Source And Urgency
09:43 How To Start Today
10:44 Get The Full System
11:20 Wrap Up

Full transcript

GLM 5.2 is a really powerful AI SEO agent, and I'm going to show you why exactly that is, plus what we built with it in about 30 minutes flat. So here's what just happened. On June the 13th, a company called Zed AI dropped a brand new model called GLM 5.2. It's got a million token context window, great for creating content, great for agentic tasks, and that's exactly what makes it so powerful for AI SEO.

And that's not why I'm excited, actually. I'm excited because what you can actually do with it. Now I've been testing it all day, and what you can actually do is you can plug GLM 5.2 into an agentic tool. So the one that we actually used was Hermes Agent, and we plugged it into GLM 5.2.

And once you do that, you don't just get one AI doing one task, you get a team of AI workers building things for you. And that could be websites, it could be blog posts, it could even be videos, even games if you wanted to, whilst you go and make a coffee. So let me prove it. We took GLM 5.2, we plugged it into Hermes, and asked it to build a website, a real one, with blog posts designed on it.

And it is super fast to load. It did it in about 30 minutes. We didn't write the blog posts, we didn't pick the colors, we didn't build the pages. The agents did all of it.

And you can see how quickly it is to load, and also how nice it actually looks. Now, the agents just handled this for you. And here's the part that really matters for you. That website wasn't just built pretty, it was built to rank, to show up when people search on Google to bring in customers without you paying for ads every single day.

And I'll show you how that ranking actually works in a second, because the proof is real, and the numbers are actually right there. But first, let me talk about probably the biggest lie that you've heard about this sort of stuff. So a lot of people think that you need to be like a coder to use AI agents. But at this point, it's so easy to just go inside a Kanban board and give it a chat, right?

This is not for just tech people in hoodies. If you can't write code, you're not locked out. And it's something that most people don't even realize. So GLM 512 lives inside a coding plan that starts at around $18 per month, one flat price.

So you don't get charged per API, you just use it. And the agent tool, Hermes, does the talking for you. So you type what you want in plain English, tell it exactly what you need, and then it goes from there. So let's say, for example, you run a cleaning business or a coffee shop or an online store or a coaching service, or anything like that, this is for you.

It's not for coders. And the thing that I keep saying is this, because people don't believe me until they see it, right? Inside the AI Profitable Boardroom, we've got over 3,600 business owners, and a lot of them had never touched AI before. They joined, now they're building websites and agent operating systems and watching them rank as well.

So if they can do it, you can do it. Now, let me show you the team part and how that works, because this is where it gets really interesting. When you plug GLM 512 into Hermes, you can make different AI workers, like hiring a small crew. So you can set up a researcher, a writer, a builder.

And here's my favorite one, a judge. The judge is a secret. Most people use AI to get like so-so results, and then they just stop. The judge doesn't stop.

It looks at the work, all the other agents did, and then tells them to fix it, do it again, make it better over and over and over again until it's actually good. And the great thing about that is that you're not babysitting the agent, the judges. And it iterates until it's actually good. So picture this, for example, you want a blog post about your service.

The researcher can dig up the facts. The writer can write it. The judge reads it and says, not good enough, redo the intro. The writer redoes it, you come back later and it's actually done.

That's a whole content team working whilst you're doing something else, or in the gym, or that sort of thing. And let me show you the website we built, because this is the proof. We had four agents working together on this, a researcher, a writer, and the judge to score it. They built a full website, multiple pages, blog posts on it.

And the site actually looks super nice, right? It actually explains what they did side by side in terms of how agent teams work and how to build them. So it kind of had this meta and it explains how they work as a team, step-by-step, so anyone can understand it. Now the design looks clean, looks professional.

It was actually nicer than Claude Opus' design. And it was fast, like really, really fast. So when you click between pages, as you can see here, it moves and it loads very, very quickly. And that speed matters because obviously Google likes fast websites as well.

So do your customers. Nobody likes waiting around for a slow page anymore. Now, was it perfect in one shot? No, we had to go back and forth a little bit, but that's the whole point of the judge.

It keeps pushing until it was right. And the whole thing was done in about 30 minutes. Here's an example. So you can see on this particular Kanban board, we've got the history and you can see the judge working over there, the SEO writer, the GLM researcher, the GLM writer, and the GLM SEO optimizer.

They're all working together as a team on a Kanban board. Now, if you're wondering, does this actually work in practice? You can see multiple websites that were built using the same method with Hermes Agent and they're all popping off, right? You can see the trajectory here.

They're all moving in the right direction. They're all getting great results. There's four different case studies we've got right now, all of them ranking and doing very well when it comes to SEO using teams of agents, which means real visitors, free traffic, and a slow, steady climb that is bringing in leads. And the great thing about a ranking website is it's working 24 hours a day for you, right?

It's bringing people in day to day. And now you can build with those teams using GLM 5.2 on the agent plan. So if we drop in a task here for SEO, it's just going to go off and complete it. By the way, if you'd like this sort of stuff, we actually have a free SEO course of over 260 lessons on how to do SEO.

Link in the comments and description. And if you want a free SEO strategy session, just go to goldie.agency and we can look at your website one-to-one. Now, we even built out a full agent operating system. So this is actually like operating system inside the OS with different apps, as you can see here.

And there's something even bigger than this that you can do with this. So you can actually make videos too. And videos rank very well with SEO. So if we have a look at this particular keyword, best AI community, you can see us ranking inside Google AI reviews.

You can also see us ranking over here. And then you can also see us ranking on the first page of Google over here too, as well as here, right? And so you can rank multiple times for that keyword, as you can see, by using videos. Now, how do you do that?

So we actually use GLM 5.2 to make this video, as you can see. And we basically had Hermes make a full video of me. So it was an AI avatar, the agents wrote the script, they made the voiceover, they picked the look, they even did the editing. I didn't touch it at all.

Not one part, I just told it what I wanted and the team built the whole video start to finish. And those videos can rank, they can show up, they can get you found, they can bring potential leads. So think about what that means for a second. If you've ever wanted to make videos, but you don't like to be on camera, or you don't have the time, or you don't know how to edit, this does it all for you without any humans involved.

So a team of agents can build your video whilst you literally go off and do something else. Let me cover the next thing people throw at me usually when it comes to this. So people say, you know, I already use Claude, why would I switch? And Claude is great, but here's the honest problem.

If you want to run Claude inside an agent team like Hermes, you have to use the API, you can't use his subscription. Whereas you can build as many agent profiles as you want using the coding plan on GLM 5.2, and it's flat on the plan. So you can run as many agent tasks as you want, it counts the same. And because it's built into your own system, you keep everything.

Every website, every blog post, every video is saved in your workspace ready for when you need it. Now, some people at this point are going to say, well, you know, I should just pick one AI and stick with that. And I get why that feels safe, but here's the truth. The models change constantly.

Something new drops every week. GLM 5.2 itself just came out. So if you lock yourself into one tool, you fall behind the moment something better shows up, or when something gets taken away, for example, like Fable 5 yesterday. So the smarter move is a system where you can swap models in and out.

If a new model drops out, you can plug this one in. If GLM 5.2 came out today, well, we've already plugged it into the agent operating system. That's how this works. So you're never stuck, you're never behind.

And that's really the old way versus the new way, isn't it? The old way is like you're stuck in one chat box with AI doing one task at a time and losing your work. You can't find it inside the terminal. If something breaks, you're kind of stuck.

The new way is you've got a whole crew of AI agents working together, and then you can manage them inside your profile section with Hermes, and they can be all in one dashboard, building websites and videos that actually help you, saved in one place, whilst you go off and do something else. And you can swap the brain behind them whenever a better one comes out. Yesterday, Kimi K217 came out, we plugged it in. Then GLM 5.2 came out, we plugged that in, right?

That's a shift, and it's happening right now, whether you're part of it or not. Now, let me show you one more thing about how this whole runs, because it's way smarter than you actually think. When you give a task to your Hermes crew, inside the Kanban board, this is like a to-do list for your team. So if we have a look at the blogs, you can see that everything was done over here.

Now, when you give a task to your crew, it picks the right worker for the job on its own. You don't have to sort it. If it's a video task, it grabs your video writer. If it's a blog post, it grabs your content writer.

It reads what each agent is good at and matches the job to the worker. So you're not just like micromanaging a bunch of AI tools. You just drop the task inside here, and then the system handles who does what. And you can watch it all happen.

Every build gets saved. So later you can go back, look at what works, use it again, nothing gets lost, nothing gets forgotten. And let me get real with you about where this is going. So GLM 5.2 is going to be open.

Next week, the company is releasing the full thing for free under an open license. That means more tools will support it. More people will build with it. If it's open for everyone, it's going to be everywhere.

And the businesses that learn this now, whilst it's still early, whilst most people are typing one question at a time into a chat box, those are the ones who pull ahead. Not because they're smarter, but because they actually started. I'm not telling you like the sky is falling. I'm telling you the door is wide open right now and it won't stay this wide open forever.

So here's what I'd do if I was you. If you've never touched this stuff, you can just start small. You can get GLM 5.2 on the coding plan, plug it into an agent tool at Hermes, ask it to build you one simple page and just see it work with your own eyes. And if you've already used AI a bit, go a bit further, right?

Build a small crew, a writer, a researcher, a judge. Have them make you a blog post. Watch the judge push them to make it better. If you run a business and you're serious about getting more SEO traffic for your website, well, this is probably the most powerful I've ever seen an agent team working together.

I've never seen it so easy to create a full team of agents, create videos and blog posts parallel, and then just go off and build amazing stuff at the same time. Yeah, I can work 24 seven for you. You can create scheduled tasks for this. You can go off and build amazing stuff in the background.

And that's the whole point of GLM 5.2, right? It's not like a clever new AI. It's really finally letting people like us with no coding skills run a team of AI workers that build real things and bring you in real customers. Now, if you want my exact setup, the full agent operating system where GLM 5.2, Hermes, Claude, and all your AI agents live in one place, come into the AI platform boardroom, you get the system as a download ready to install.

You get a 30-day roadmap that walks you through building websites and videos for GLM 5.2 agent teams and getting them to rank so they actually get you leads. You get daily tutorials, four live coaching calls a week to ask about your own business, and over 3,600 business owners build in the same way, many of them right inside GLM 5.2 already. And we update the system constantly. So as new models drop, you're always current.

Links in the comments and description or go to the AIProfitBourne.com. Hope to see you inside there. Cheers.

More episodes

Browse all episodes →