Hermes + OpenClaw: NEW AI Voice Agent is INSANE!

With the new MINIMAX M3 update, you can now build a voice agent directly into OpenCore and Hermes. Let me show you exactly how this works. So if we click this, we can start the voice chat and speak to our agent directly. Hey, how's it going?

Well, ready to help you with whatever you need today. What do you think we could automate that would be quick and easy related to SEO? Weekly rank tracking and broken link checks with a simple script or tool like Screaming Frog. What's the funniest joke you know?

You told my computer I needed a break and it said, no problem, I'll go to sleep mode. What's the most interesting fact you can tell me right now? Teach me some Swahili. So you get the point.

Basically, you can build out a voice agent into your Hermes and your OpenCore. We can do the same inside Hermes here. You can see we just tap to it and then speak. And then there's multiple different voices.

You're welcome. Let me know if you want to learn more. And then there's multiple different voices that we can plug in here as well. I don't know why I kept going there, but there we go.

Now, this is actually based on a question from Paul inside the AI Profit Boarding. He was asking like, you know, how can we speak to OpenCore? How can we have a chat with your AI agents? How can you speak to them in real time, et cetera?

And this is a powerful way to do it. So essentially what we're doing here is using a framework I call Voice Agent Mastery. And this means that you stop typing to your AI and you can talk to it. So, you know, imagine picking up a phone, saying what you need out loud, and your AI answering back in a real voice, right?

So you don't need to type. You don't need to click. It's hands-free. And that's a voice agent.

And it can run inside an agent operating system like I've just shown you. So you can speak. It hears you. Minimax thinks.

Minimax talks back. Turn after turn. Like a real conversation. And so this means, number one, you can interact faster with your AI agents.

And also, like, let's say, for example, you're on the go. You don't have time to type. This is a faster way to just go back and forth with them and actually operate. It's really the difference between, like, texting your AI, which is, you know, just using your keyboard and then actually calling your AI agent as well.

And the reason this works is because basically you're interacting with Minimax M3. Minimax M3 is the most agentic model I've used with Hermes agents so far. It's really fun to use. Easy.

Super smooth. It can also do a bunch of other cool stuff. So, for example, if we go inside the studio here, you can see that we can generate images. We can generate videos.

We can generate voice, etc. And the same inside OpenClaw. We can generate voice, video. We can talk to it.

We can use images. You could also use Grok as a voice as well. So, for example, if you use Grok and you plug this in too, then you can actually use Twitter search and you can search Twitter in real time, which is pretty cool because if you've got AI agents and you're, for example, like researching the latest AI automation news each day, well, you could put that on a schedule and get that running directly too. You've got four different voice modes as well.

So if we test out the Hermes one here, let's try the deep mode. Hello there. Are you listening? Yep.

I'm hearing, listening. Go ahead. What's on your mind? So that's pretty cool.

And then also when you're doing this, you can also save voice notes inside your workspace. So, for example, if you were generating like podcasts, you wanted a place to store them and come back to them later, well, you could have them inside the workspace and then you can switch between like the studio, the talk mode and the workspace. And that's pretty powerful too. So the interesting thing here as well is like you could operate this from your phone.

You know, you could have Hermes agent inside Telegram, maybe you're at the gym or whatever, and you can come up with new ideas hands-free whilst you're away from your desk. Also, something to note here is like it's not 100% instant. So just be aware that there might be a couple of seconds in between like when it transcribes and when it thinks and then when it actually comes back to you and speaks. So just something to bear in mind there, but it is pretty powerful stuff.

And so the way that this works is basically you speak into the microphone, it turns the voice into text, at minimax M3 thinks, and then it speaks back to you. So it's pretty powerful as a workflow. Now, some people say, well, talking to AI is a gimmick, typing is faster anyway. But, you know, if you're doing something or you're out and about, or for example, you're not that fast at typing, or for example, it's quite energy demanding typing, then you can talk to your AI agent through the voice instead.

And also some people say, well, voice assistants never understand me. But you've seen, for example, today they actually work perfectly first time round every time, which is great as well. On top of that, you might think you need special hardware or something like that, but actually you can just use a combination of minimax and either OpenClaw or Hermes agent. And then you might say, well, this sounds complicated, but it's already built into this system.

So you can use this directly inside the agent operating system, have a conversation with your AI agents, and then everything that you create can be stored inside the workspace as well. Along with all the other projects you've built, I mean, there's so much cool stuff you can build with minimax. Like look at this website that we built. Here's another example of a blog post we created.

You can see some examples of voice agents we built here. Like there's all sorts of cool stuff you can do with this and it can generate images and videos and everything else. So that's basically it. That's how to use it.

Just to recap there, you no longer have to type at your AI agents. You can talk to them hands-free and it'll talk back to you in a real voice. It works because it's using your speech, transcribing it, and then minimax will think and come back to you. We've plugged it into the agent operating system if you want to get it.

And that works with Hermes agent and OpenClaw. The other thing I would say here is if you haven't tried out minimax M3, it's pretty good on the coding plan. The reason for that is because, you know, if you're using the API, it can use up a lot of tokens very quickly. Whereas for example, if you're on the coding plan, I haven't had any problems like that.

And it's designed for agentic capabilities. It's actually one of the only tools I've seen that's run for like 24 hours autonomously on their tests. And it works pretty well with kernel optimization, which basically means they gave it a hard coding task. It ran on its own.

It tried over and over again and did like 1,959 agentic tool calls and then came back and actually completed the task and did it very well as well. So it's a pretty powerful tool. And again, like because it's minimax, it can generate images and videos using a combination of minimax and all the other models around it. So thanks so much for watching.

If you want to get more trainings and stuff like this, you can get it inside the Aircraft Reporting. If you want the agent operating system that I've just shown you, you can get it over here along with the zip file. We actually update it daily. So with new updates that come out, we've got those plugged in.

You get new video tutorials every single day. You can connect with me personally. You can ask questions inside the community and I'll answer them with video tutorials like I did today for Paul. And then also inside the classroom too, you can get access to all my training.

So if you're like a complete beginner, you can go from beginner to expert with this training and we've got loads of cool stuff as well. Like if, for example, you're trying to build a memory into your agents, you can do that using this Obsidian training over here. On top of that, we have four weekly coaching calls where you can ask questions to get help support in real time. You can connect with people inside the map and meet people in your local area using open core AI agents like Hermes, et cetera.

And that's all inside the AI Profitable Boardroom. Link in the comments description or just go to the AIprofitableboardroom.com. Thanks for watching.