Hermes + WebUI Just Changed AI Agents Forever

Today we're going to be setting up Hermes web with Hermes agent. So Hermes is a sophisticated, autonomous AI agent, and you can link this to web UI. Now, one of the benefits of this is like, it looks a lot nicer than the terminal user interface. So if we load these up, I'll show you an example and compare the differences.

So this is an example of what the terminal user interface looks like for Hermes. And it's okay, but it's, it's difficult to manage sometimes, right? Whereas you can see an example of what the new web UI looks like. And this is a free way to use it, right?

So this is Hermes web UI, and you can also access it from your phone as well, which is pretty cool. But this is what it looks like once it's set up. Now, if you want to get this set up, there's a couple of ways, right? Number one, you could run this directly here, or you can run it inside Docker as well, right?

So let's have a look at an example of how to quick start this. So we're going to copy this command right here. The other way you could do it is you could actually ask Hermes to set up directly for you. If you don't like using the terminal directly, but you can see how easy that was to set up, right?

So we literally just ran that command, boom. We've now got this set up right here, right? Now we can update this. So you can see that we've got a bunch of updates to set up.

So we can do that right now. We've got Hermes as you now look how much nicer the of this looks versus this, right? It's going to be much easier to organize. It's going to be much easier to speak to the AI agents directly, and you can also manage multiple other agents.

So you could, for example, have Claude working over there. I think you could link this to other tools too, but this is specifically designed for Hermes UI, right? So let's try this out now, just check it works as well. So we've got it connected to our agent there.

You can see the green online section, which means that this is actually online. You can also switch between providers, which is pretty cool. So you've got a bunch of different models that you can switch between, which is useful too. And then you can select between the reasoning modes here.

So you've got minimal, low, medium, high, but bear in mind, like this is just a free UI. You can set up in a couple of minutes with your Hermes agent. If we zoom in there, that looks a lot nicer now, right? So let's try this now.

We're going to say, okay, you're working just to give it a quick test. Bear in mind, it may be set up, but it might not be connected. So it's just something you want to be careful of. We'll see whether it responds right now.

If you have any questions as well, if you're watching live, like feel free to ask as we go along. And you can see here, for example, this actually struggled to answer. So that's what I mean. Like sometimes you can set these things up, but they're not actually linked directly to your AI agents.

Particularly if you're using a model that maybe the API's key is changed on it or something like that. Right. So I'm just going to search models here and we'll find the setup. I think we might be able to change configuration as well.

Yeah. If we go to providers here, we can switch between providers, right? So we've configured that, but I would much rather use something like open router. So I'm just going to grab another open router setup, which we can do over here.

You could also run this with open code as well. Open code is a good option. ZAI Mistral. Mistral is a local model you could use.

And DeepSeek as well. So they've got DeepSeek here as well. So you could link this to like DeepSeek v4 or something like that. Spark says, in your opinion, how does Hermes Agent compare to Cloud Code?

So for me personally, I find Cloud Code is much easier to use. It's a lot less technical and it's easier to manage. It's kind of like the difference between building your own car and getting one that's ready to drive straight away, right? With Cloud Code, it's ready to drive straight away.

There's very little configuration, but you don't customize it as much, right? Whereas for example, with Hermes, like you can give it its own personality. You can customize it a lot more. You get these amazing open source projects like this.

And so that's the biggest difference. And also of course, like you get more control over the model that you're using. So for example, if you're using Hermes, you can switch between all these models. No problem, right?

But obviously if you're using Cloud, then you're using Cloud models as well. And that means like if they change the plans at any time, which they have just recently, then sometimes you can lose access to features and stuff as well. So I'm actually going to remove Minimax because that didn't work. So we'll remove these over here and then we can configure Olama as well.

So what we can do over here, I think this might be the best way actually. I'm tempted to use Olama, but let's try OpenReta first. So we're going to go to settings here, API keys. We'll grab an API, create, and then we'll go over to here, paste that in, hit save.

And that should be configured now. So now if we go back over here, we should be able to switch to OpenReta. So you can see OpenReta is working there. And then let's just try this out.

Nice. All right. So that is working now, right? Um, for some reason, it switched back to Minimax M2.7 via the Olama setup, but either way it is working fine.

And we're just going to check. Are you Hermes? Now, if we actually open up the terminal over here, it's not interacting with the terminal session that we have. Just bear that in mind.

And if we have any issues setting this up, then what we can do is go back over to Hermes and terminal and just ask it for help on this, right? As you can see here, it's, it's, it's a little bit buggy. I think that's because it switched the API here, right? So the model takes change in your next conversation.

So if we go on a new conversation here and we just need to, yeah, that's working now. Yes. And it says, yes, I'm Hermes, build by Noose Research, blah, blah, blah. So now we can actually have this really nice UI for managing our agent, right?

And just look at the difference here, right? That's my terminal for Hermes, right? Super messy. That's a web UI for Hermes speaking to it directly, right?

How much nicer is, is one versus the other. So that's the biggest difference here. If we go on the left-hand side, it's also responsive for mobile, which is great. And then we can easily reconfigure the models.

We can add a to-do list here. So we're going to add a task list. If we have an active session with a task list, we can change our profiles, right? So we can add a new profile here, add the URL as well.

And the API key to link it to different profiles. We can actually select a space to switch between sessions, which is pretty cool. And then we have personal memory over here, right? So you can pick a section.

So for example, we can pick between these two notes here, right? The same with the sessions over here. So you can see that we can select which space we want and we can also add new spaces too, if we want to add new spaces. So this is pretty good because it now has like all this, all these details on my memory, we can manage that actively.

We can manage our spaces easily. We can manage our different profiles, which is great. And then also we have the scheduled jobs, right? So these are the daily scheduled jobs and the tasks that are going to run automatically.

So if we wanted to delete one, like let's say, for example, we've got kind of a duplicate going on here. We've got one with an error and one that's active. Well, what we could actually do is delete that one just to manage it better and clean it up. And then over on this section here, we can manage our daily active tasks, right?

Now, if we refresh, we add new ones inside Hermes. We can just refresh and change over here and then we can edit the prompts directly. Right? So we can go to edit and then we can change the prompt.

If we want to edit this prompt, we can change where the output actually goes to. So whether it's cloud or whether it's to local, et cetera, we can change the schedule, so how often it runs and when it runs and what time, and then we can also add skills, right? So if we X off that, that's how you can manage your skills. We can delete them if we need to, which is pretty easy and simple.

And then over here, what you're going to find is that you have all these different skills for your AI agent, right? And so these are all the skills that you have set up directly with Hermes, which is great because you can easily manage them, you can set them up, you can change things, you can customize it, and so you just get a lot more control. It's easier to manage, it's easier to organize your agent. And also it's way nicer than the UI of the terminal.

And I think until Hermes, you know, eventually release an update where you can do this directly inside the dashboard, this is maybe one of the best ways to do it. It actually seems way more responsive and easier to use than open web UI, which is another open source project. So if you haven't seen that open web UI is another way that you can manage your AI agents. Here's an example of how that works.

I've got another tutorial on it as well, but I think the web UI just looks a lot nicer and it feels a lot more responsive and it feels easier to set up as well. Let's see what we got on the questions here. Thanks for the content, most consistent. Oh, thank you very much.

Appreciate that. This true future, agentic AI co-generator 100%. Do they have a unified chat yet? No, I've not seen that anywhere.

I've not seen like a unified chat. Um, I think the, the closest I've seen to that, I think is Aon web UI. Yeah. Aon UI, right.

So Aon UI unifies all your agents in one place. Kind of similar to a paperclip, for example. And you can, for example, like manage all your agents in one place. So you could have like open core, you can have core code, you can have Gemini, Quan, Hermes directly inside there as well, as far as I'm aware.

Yeah. You've got Hermes inside there too. So that's like one way you can unify it, but I'm not seeing anywhere where you can unify all the messages across all your channels, like, like I was saying, for example, if we use open web UI or web UI itself, it's not going to sync directly into the terminal. Right.

There might be some history. There might be a place where you can see like the history of all the messages and interactions kind of like you can with open core, but I've not seen that and maybe I'm just not aware of it and that's it. So the open core has it built in the standard gateway. Yeah, that's it.

But you, you don't get that with Hermes as far as I'm aware. Um, however, I, you know, there's, there is one way that I know if you can do this. So what you can do is you can do Hermes dashboard inside your terminal, as long as you're upgraded to version v0.11. And what that's going to do is build a web UI here.

And so once that opens up, you can actually see your previous sessions here. Right. So you can see all your, all your sessions here. So you can actually see, zoom in here.

It's actually got the sessions where I've asked, like, are you Hermes? And you can see where it's done. So I think that's probably the easiest way to manage all your sessions in one place. Right.

So you can see, for example, we've got, are you Hermes? Are you working? Right. That is me testing out web UI right here.

And then we have, for example, the system schedule tasks, which we'll work in there. And then we also, for example, we have the workflows I created yesterday on AI SEO. And I think that might be the best way to manage it actually. Cause you've got it all inside one session.

It's easy to organize everything synced in one place, but bear in mind, like you can't talk directly to this so you can see where it was. Like this says web UI, web UI, schedule task, CLI. So you can see all your sessions and all the places you've talked, but you can't like directly message the AI. We, so it's kind of one of those, like, I I'm kind of surprised that news research haven't built their own sort of messaging section inside the dashboard.

But I assume that's probably coming. Bear in mind, we're not even on like version 1.0 yet. What is terminal? Terminal is just a way that you can talk to your agent and organize it right locally.

So it's, it's just kind of like a way that you can run local commands on your computer to manage your agent. I just, I just want to be able to message it from Telegram, pick up my, yeah, well you can now you can use the open core mission control I just showed you a minute ago, right? And the way that you open that up is you just run open core dashboard and called dashboard inside Hermes. It will build out there and then you can see all your sessions, right?

That's how you can use it. So if we go back to this page, boom, look at that, we can see everything that's running right web UI and everything. So that's the easiest way to manage it. But if you want to just chat directly inside, like a kind of nice, simple user interface, and you can see all your conversation history as well, which is great, you can filter conversations, manage the scheduled tasks, manage skills, manage memory, manage the active folders, everything else that's all right here.

So that I think that's one of the best ways. Bear in mind as well, this is super useful. So like, if you've never done this before, you can create multiple profiles, right? You can create multiple profiles for your AI agents.

So you can see, I've got John, I've got Julian. Um, I've got the default and I've got this one as well, right? And so these are all different versions of my Hermes with different skills. As you can see, this has 88 skills.

This one has 155 skills, right? So they all have different skills attached to them. This is the only one that's active right now. And so it's super easy to just manage my profiles.

If I wanted to do that inside the terminal UI, it's messy. It's very messy, you know, so fun. I'm running it right now with DeepSeek v4, me too. I'm running it with DeepSeek v4 too.

It's awesome on that. Sweet. Happy to help. You're clearly the best in this space.

Thank you. Oh, I appreciate that. That's super kind. Appreciate that.

I don't know if I am. I don't, I actually don't think I am, but, um, I really appreciate that. That's super kind. This is open web UI.

No, this is web UI, different projects, right? So there's open web UI. Let me pull this up. So you've got two different, two totally different projects here.

You have open web UI, which is a GitHub project. And that works across like all your AI agents. So it's not just designed for Hermes, right? Then you've got Hermes web UI.

And Hermes web UI is purely designed for one thing, speaking with Hermes. And so these are two totally different open source projects. You can get them both for free and they're both pretty good. But if you just want one single place for Hermes, then you would use Hermes web UI, right?

And the good thing about that as well is like, you don't get confused with your other agents. If you're just focusing on Hermes, this might be the best way to do it. Um, what plan do I suggest? Well, I mean, you can go with something like a minimax.

Minimax is, is pretty affordable. Um, Kimmy as well. Kimmy is a good option. Web UI is better.

Web UI is better so far from what's tested. Going to test Hermes web UI. Yeah. Um, yeah, I, I do have the Olama premium plan and it is so good.

It's just so easy. You know, like if you, if you never, someone asked me like, you know, which model am I using? So if you've not, if you're not familiar with Olama, basically for every single model, you can have Olama running locally. And then if you just want to run open core with it, then copy that into your terminal, right?

You want to run Hermes with Olama and deep seek, boom, copy that into your terminal, right? And the, one of the best things about that is number one, you can set up the agents in one single click. And then number two, you can easily, um, switch between models, right? You can see all the different models here.

So as long as, so, you know, as soon as something interesting comes out, it's right there. And the good thing about this is if we go to the settings here, I mean, number one, you can try it for free, but number two, even if you're on like a, you know, a plan like this, for example, I'm on the pro plan session usage, like I use this every day. We've barely used it, right. Cause it resets every like 29 minutes and then we have a weekly usage, but you can see I'm far from using that.

So yeah, Olama with the pro plan. Awesome. Awesome stuff with the pro version. You add that.

Yeah. And also you get more usage, right? That's what we like. Am I planning to talk about the space agent?

Not today, but I have lots of training inside the AI platform on that. If you don't code as like a subject and just need the general AI, is Kimmy good? Um, I would go with Claude. Uh, all of those I go with Claude.

It's the best. I'm in the process of setting up paperclip, Olama and Hermes. Yeah. It's really good there.

Yeah. I've got training on space agent already. If you haven't seen that, how to create AI agent platform. So you can just run, I mean, like for example, this, this is kind of like a So you can just run this with Hermes and it was a one click setup, right?

It was literally just like copy and paste the quick start and you're good to go. So that is basically how to use Hermes web UI, how easy it works, how to set up, how to sync it, how to manage your profiles, your skills, your memory folders, everything else, pretty simple and easy. And, um, if you want more training on this sort of stuff, feel free to get the AI profit boardroom link in the comments description, this is my AI automation community where I cover stuff like AI agents, how to set them up, how to use them, how to get the most out of them, how to automate your business with them. We have new daily tutorials on this stuff, right?

And we have a full two hour course on Hermes, we have a full six hour course on OpenClaw and we add new interesting tutorials every single day, right? As you can see, um, including for example, like Hermes and open web UI, plus you can ask questions inside the community. There's loads of people using this stuff to really master Hermes and AI agents like OpenClaw too, plus inside the map, you can jump on weekly coaching calls, meet people inside your, your local city who are using this stuff too. So feel free to get that and you can connect with me personally inside that too.