Hermes Agent: The Ultimate Beginner’s Guide

Hey, I'm Caleb, and in this video, I'll teach you exactly how to set up your own Hermes agent and everything you need to know to actually use it in the most beginner-friendly way possible. By the end of this video, [music] you'll have your own self-improving AI assistant running 24/7, messaging you on Telegram, learning from every task you give it, and handling stuff on a schedule so you don't have to. So, here's what we'll do next. First, we'll deploy Hermes on a VPS using a one-click template that handles the installation for you. >> [music] >> Then, we'll connect it to an AI model provider and set up Telegram so you can chat with your agent from your phone. After that, I'll walk you through everything you can actually do with it, how skills work, how memory works, how to set up automated jobs, manage costs, and a few other things that Hermes can do for you. Okay, so first, let's talk about what Hermes agent actually is. So, Hermes is a self-improving AI assistant. It runs 24/7, connects to tools you already use, that's stuff like Telegram, Discord, and Slack, and it gets better the more you use it. Instead of living in a browser tab like ChatGPT or Claude, it runs on its own server. It can also stay online around the clock, and it can message you first when something important happens. And you can use Hermes for all sorts of stuff. That could be daily briefings, GitHub help, trend monitoring, trip planning, and a whole lot more. Basically, whatever you need help with. So, what really makes Hermes different comes down to four things. First, it can write its own skills. When you give it a task that takes a real work, it's able to save how it handled that task and create a custom skill based on the exact process that it used. So, the next time you ask for something similar, it already knows exactly what it needs to do. Second, it remembers you properly, not just random details, but how you work, what you prefer, and how you like things done. So, your conversations with Hermes will build over time instead of resetting every time. Third, you can schedule tasks using normal language. You could say something like, "Every Monday at 9:00 a.m., send me a summary of trending news in my niche," and it just does it automatically. And fourth, it works across more than 15 different messaging platforms through a single gateway. So, whether you're using Telegram, Discord, Slack, WhatsApp, or Signal, it's still the same agent with the same memory. It doesn't matter where you're talking to it. So, you have a few different ways to run Hermes. You can run it on your own computer, on a spare mini PC, or on a VPS, which is basically just a separate computer in the cloud. For this video, I'm using a VPS, and that's the option that I recommend for most people. The biggest reason is that a VPS stays online 24/7, and that really matters because a lot of what makes Hermes useful, like scheduled jobs, alerts, and the always-on gateway, only work if the agent is running all the time. If you install it on your laptop, for example, it's going to stop working the moment your computer goes to sleep. A VPS also gives you a clean isolated setup, which is great for security. And if something ever breaks or you just want to start over, you can easily wipe your VPS and redeploy all without affecting your main computer. So, for this setup, I'll be using Hostinger. They have a one-click template for Hermes, so you don't have to install everything from scratch. You just fill out a few details, and the setup handles the rest. To get started, go to this link on screen right here or click the first link in the video description. Using that link will make sure you get the lowest pricing available. And once you go to that link, you'll be brought to this page right here. Start by going over to the right-hand side. Here, you can choose your VPS plan. By default, this will be set to KVM-2, and this is the plan that I recommend for running the Hermes agent. It's going to give you two virtual CPU cores, 8 gigs of RAM, 100 GB of NVMe disk space, and 8 TB of bandwidth. This will be plenty of resources to get you started. So, go down and click on deploy. From here, start by choosing your billing period. By default, this will be set to 24 months, which is a pretty long-term commitment. You can switch this from 24 months to 12 months and 1-month plans as well. I'd recommend going with the 12-month plan. This will still get you plenty of savings and a free domain for the first year. Scrolling down, you can choose if you want to toggle on daily auto backups, and you can also choose your server location. Hostinger will automatically choose the server location with the best latency for you. So, I'll leave this as the default location. Let's scroll up to the top. You can review your order here on the right, and don't forget to use the first link in the video description to get an additional discount here. With that done, go down and click on continue. From here, go ahead and enter your email address, create a password, and click on the register button to move on. Then, on this screen, fill out your billing address and your payment information, and I'll meet you at the next step. From here, click on get started to move on. Okay, and once you get to this screen right here, you'll need to set an admin username and an admin password. This is just for logging into the Hermes web terminal later. For the username, the default is Hermes, and you can just leave it as that. And for the password, Hostinger will automatically generate a strong password for you. I'd recommend leaving the default password as well. I'll go ahead and reveal this, copy it, and I'm going to save the username and password somewhere safe on my computer so I can reference it later. All right, I've gone ahead and saved the username and password. Once you've got those saved, go down and click on deploy. Scroll down here and click skip, and you'll see that our Hermes agent project is being deployed. This process can take a few minutes to complete, so I'll fast-forward to whenever it's done. Okay, and now that our project is deployed, this is our VPS dashboard. You'll be able to see all of your server info and all of your different tabs and settings here on the left. And right here, you can see your Hermes agent listed as a Docker project. Let's go ahead and note down the last four letters right here at the end. I'm going to copy this and save it somewhere on my computer to reference in just a few moments. Okay, and with that done, let's go to the top right and open up the terminal. And right here is our Hostinger VPS web terminal. It's what we use to communicate with our VPS directly. And if you have never used one of these before, don't worry. They can look a little bit intimidating, but I'll walk you through exactly what you need to do here. And I also want to let you know that if you check in the video description, you'll find a link to a Hermes agent setup guide. It's going to list out all of the different commands that we use throughout this video. It's going to tell you what they do. It's also going to list out some other helpful commands. So, go ahead and check the video description right now and go to the Hermes agent setup guide. You can have it open in another tab or on a separate monitor, and it's going to help you with following the rest of this video. Now, the first thing that we want to do here is navigate to our project directory. Let's go ahead and run this command. CD /docker /hermes -agent and then enter the last four characters that we just saved a moment ago. And then by using control, shift, and V on my keyboard, I'll paste in those last four characters right here at the end of the command, and then press enter. All right, now let's get inside the running container. First, type docker compose exec -ec -it hermes-agent /bin /bash Press enter. And now we're inside the Hermes agent container. Now, we're going to run the Hermes setup wizard. This is what's going to walk us through connecting an AI model and getting Telegram set up. So, let's type hermes setup then press enter. Next, you'll see this question right here. It's asking you how you would like to set up Hermes. For this video, we'll be going with quick setup. So, with quick setup selected, press enter on your keyboard. Now, you'll see a list of all the AI providers that Hermes works with. There are a lot of options right here, and you can check any of them out, but for this video, we'll be going with OpenRouter. OpenRouter gives you access to over 200 AI models with a single API key. So, instead of being locked into one provider, you can switch models anytime with one command. So, on my keyboard, I'm going to press the down arrow, and with OpenRouter selected, I'll press enter on my keyboard. And then below this, you'll see that there is currently no OpenRouter API key configured. It says we can get one right here. So, let's highlight this link, then we'll right-click on it, and we'll go to this link right here. And here, if this is your first time using OpenRouter, you'll need to sign up for an account. So, go down and click on sign up. Then, go ahead and enter your email address and create a password. Check this box right here and click continue. And here, it says it's going to send a verification link to your email address. So, go ahead and check the email address listed right here. You'll see an email that looks like this. Scroll down and click on sign up to OpenRouter. For this here, I'll click other and continue. Then, click create API key. Click create. For the name, we'll name this Hermes agent. And here, you can choose a credit limit. If you leave this blank, it's going to be set to unlimited. So, if you want a cost guardrail and you're not looking to go over a certain amount, I would go ahead and enter your limit right here. In my case, for my agent, I'll set this to, we'll say, $50 for now. And then here, you can choose when you want to reset your limit. You can and between daily, weekly, or monthly. For now, I'm just going to leave this turned off. And at the bottom right here, you can choose when you want this API key to expire. You'll see a few options right here, but again, for this video, I'll just go with no expiration. All right, with that done, go down and click on create. And right here, you're going to see your API key. Go ahead and click on this copy button right here. And make sure that you don't share this with anybody. I'm only showing you my keys throughout this video because afterwards, I plan on deleting everything. Since you've seen my API key here, that means it's now compromised. So, be careful with your keys and make sure you don't share them with anybody. So, now we have this key copied down, but before we switch back over to Hostinger, we need to make sure that we have our billing set up. If our API key is unable to load in new credits, that means nothing is going to work. So, let's close out of this. And then right here, you'll actually see your API key. You'll also see the usage, and then your settings regarding your limit. To get billing set up, go to the left-hand side here, then go down and click on credits. And then right here, you can add credits, or you can add a payment method to automatically top up your keys. In my case, I'll just click on add credits. And then from here, I'm going to fill out my billing address and my payment information. So, go ahead and do that and go through the next steps here, and whenever you're done, I'll meet you at the next step. All right, once you have your billing set up, let's go ahead and switch back over to the Hostinger web terminal. I'll go to the top right here. And then right here, I'm going to paste in my API key. And you can paste that in by using control, shift, and V on your keyboard. Whenever you paste in the key, you're going to notice that nothing actually appears on your screen, and that is totally normal. Your key was pasted in, but it just doesn't show anything for security reasons. So, once you have your key pasted in, press enter on your keyboard. And then from here, you'll need to choose your AI model. By default, it'll be set to Claude Opus 4.6, which is super expensive. You're going to see a huge list of different AI models right here. You can use your arrow keys to go through and choose the one that you want. In my case, I'll go with GPT 5.4 mini. So, I'll press enter on my keyboard to select it. Then here, we'll see an option to connect a messaging platform. This is where we'll actually be able to set up Telegram. So, I'll press enter on my keyboard to set up messaging now. And then right here, you'll need to select the platforms that you want to configure. In this video, I'll just be walking through Telegram, but in this list, you'll see a bunch of other options like Discord, Slack, Signal, and a bunch of others. I'm going to press one on my keyboard to select Telegram. And press enter. Then press enter on my keyboard again. And right here, it says that we need to create a bot via BotFather and get a Telegram bot token. To complete this next step, you can either do it on your desktop device or your mobile device. In this case, I'll be using the Telegram desktop app. So, I'll go ahead and open that up on my computer. And then here at the top, let's search for @BotFather. And you're looking for this one right here with the verified check mark. Go ahead and select this. Then go down and click on start. Go up and click on new bot. And then you'll need to give your bot a name. I'll go with Hermes Agent. Then press enter. And the next thing you need to do is choose a username for your bot. I'll go with Caleb's Hermes Agent Bot. Then press enter. And if that username is available, you'll get this message right here. Right here, we're looking for the Telegram bot token. So, go ahead and click on this to copy it. And then switch back over to your web terminal. Then we'll paste in that key right here, again using control, shift, and V. And once again, you're going to notice that nothing actually appears on your screen. Again, this is just a security feature. Once you've pasted in your Telegram bot token, press enter on your keyboard. The next thing it's going to ask for is allowed user IDs. Right here, it says that you need to message the user info bot on Telegram to get your numeric Telegram ID. So, let's go ahead and switch back over to Telegram. Then at the top here, search for @userinfobot. And you're looking for this one right here. Go ahead and click on it. Then go down and click start. And this numeric ID right here is what you're looking for. Go ahead and click on it to copy. And then switch back over to your web terminal. Paste in that user ID right here. And press enter. The next thing it'll ask you is if you want to use your user ID as the home channel for your Hermes agent. This is where Hermes will deliver cron job results, cross-platform messages, and notifications. I'll press Y on my keyboard. And press enter. All right, and the setup wizard is now done. You'll see a summary right here of everything that just got set up. The first thing you'll see is a tool availability summary. Some things here will be marked with a green check mark and others with a red X. The green check mark means they're ready to go, and the X usually means that they need an extra API key to function. If you're wanting any of these tools to work properly, run this command right here. This is Hermes setup tools. And it's going to walk you through how to set up all of these different tools. For this video, we're not going to worry about these extra tools right now, but this is good to know in the future. And below this, you'll see where all of your files live. You see your settings, your API keys, and your data. You don't need to touch any of these manually, but it's good to know where everything is stored just in case you ever need it. And right here, you'll see a list of helpful commands. If you ever want to change your model, change your messaging setup, or run the wizard again, you can do all of that from here. And then below this, we can see that we're ready to go. And the terminal is asking us if we want to launch Hermes chat now. Let's type Y on our keyboard. And press enter. And here we go. This is the Hermes chat interface running directly in our terminal. On the left here, you'll see your AI model, your working directory, and your session info. Then up at the top right here, you'll see your Hermes agent version number, your available tools, and available skills. Right now, we have 28 tools and 75 skills available. We'll come back to skills in just a bit, and I'll show you what these actually do. Now, before we go any further, let's talk about keeping your agent safe. The good news is that Hermes is pretty locked down out of the box. There are four things that will keep it safe, and if you understand these, you can leave it running without any worry. First is the pairing code system that we just used. Only approved users can talk to your agent. If anyone else finds your bot on Telegram, they'll be ignored unless you add them. Second is container isolation. Because we're running this inside Docker on Hostinger, even if something weird happens, it can't touch your host system. That's a built-in guardrail that you get for free with the one-click template. Third is dangerous command approvals. If your agent ever tries to do something risky like delete a bunch of files or make a big system change, it's going to stop and ask you first. Make sure you always read these before you approve. And fourth is the skill scanner. When you install new skills later, Hermes will automatically check them for sketchy code before they install. And definitely make sure you're careful with installing new skills. Make sure you understand exactly what they do and where they're coming from before you install. Now, my advice for getting started from here is to just stick with Telegram for the first few days. Pick one or two skills that solve a real problem and start there. Make sure you don't connect your main email, your banking info, or stuff like a password manager right away. And honestly, I would recommend using exclusive accounts for any AI assistant you create. The main thing here is that you don't have to do everything in the first hour. Just start small, build trust with it, and expand from there. All right, now we're back here in the terminal itself. However, we actually want to chat with our agent using Telegram. That's why we connected everything earlier. So, let me show you how to do that next. Press control C on your we need to start the gateway. The gateway is what actually connects Hermes to Telegram and keeps your bot online. Go ahead and run this command next. Hermes gateway install. Then press enter. And to run the gateway, go ahead and type Hermes gateway run. Then press enter. And right here, you'll see that your Hermes gateway is now starting. And then from here, you can go ahead and close out of this web terminal and communicate with your Hermes agent directly through Telegram. You'll only use the web terminal for stuff like troubleshooting, running the setup wizard again, or installing specific tools and skills. So now, I'm going to go ahead and go back over to the Telegram desktop app. Then I'll click back into the BotFather conversation. And click on this link right here. Go down and click start. All right, so the first thing that you may see here is this unknown command message. When you click start on a Telegram bot, Telegram automatically sends /start as a command to the bot, and Hermes doesn't actually have a built-in /start command, so your Hermes agent is just letting you know that it didn't recognize it. This is totally normal, and you can just ignore it. Now, let's go ahead and send a real first message and actually start using our agent. I'm going to send this message first. Here, I'm going to say, "Hey, you're running on a Hostinger VPS inside a Docker container. Can you check your environment and tell me what tools and skills you currently have available?" The reason I'm sending this specific message as the first one is because it gives your agent context about where it's actually running. That way it knows what kind of environment it's in and it gives you a good summary of what you've got to work with right out of the gate. This is a good habit for any first message with a new agent. So, let's go ahead and send this. Here at the top, you'll see that your Hermes agent is typing, which shows us that everything should be working properly. We'll go ahead and wait for the message to appear down here, then I'll scroll up to the top here. It'll tell you here what it found in its environment, the tools currently available, the skills available as well, and this matches exactly what we saw in the web terminal a moment ago. All right, now let's get into what Hermes can actually do for you. Let's first talk about skills. A skill is basically a little recipe your agent writes for itself on how to handle a certain kind of task. Once it figures out how to do something, it saves that recipe, and next time you ask for something similar, it pulls it up and does it even better. That's the basic Hermes learning loop. So, you've already seen that we have 75 total skills available. These are the pre-installed skills that come with your Hermes agent by default. If you want to see exactly what skills these are, go ahead and type "Tell me each of the 75 skills that you currently have installed." And we'll see what it comes back with. And right here, it gives us a super long list of every single skill that's currently installed. These pre-installed skills cover almost everything that you'll need out of the box. But, if you find that you need a skill for something specific, just ask your agent what it recommends, and it'll walk you through the installation step by step. That's kind of the whole idea with these AI agents. Whether it's installing new skills, creating jobs, or troubleshooting something, you just talk to them like a regular person and they figure it out for you. So, for example here, if I know that I want my Hermes agent to help me out with stuff like video research and planning new video ideas, I can actually ask it to help me find the right skills for that task. So, here I'll type, "I want you to help me with video research and planning. What new skills can I install so you can help me with this?" In your case, just replace video research with whatever you're trying to do, whether that's setting up a daily briefing from your Google Calendar, connecting to GitHub, or anything else. Your Hermes agent will then check its environment and make sure it has everything it needs to get the job done. So, I'll go ahead and send that. So, right here you can see that my agent is actually looking through its skills and making sure that it has everything that it needs. And now it sent me another message and told me that I actually already have everything I need for this task. In your case, depending on what you ask, it might tell you the same thing, which is great, and that's why it comes with those pre-installed skills. But, if it finds a better skill or can't do what you're asking out of the box, it'll give you clear next steps on how to proceed. So, just go ahead and follow the next steps from your agent here. Now, the next thing that I want to show you here is that your Hermes agent can actually create new skills on its own automatically. So, for example here, I'm going to type, "Research the top trending AI tools right now and come back with the top three that would make for an interesting tutorial video. Once you're finished, create a new skill based on your approach and call it YouTube video research." So, what I'm doing here is having Hermes research a topic and then save that entire approach as a brand new skill. That way the next time I want to do this, I don't have to start from scratch. It already knows exactly how I like it done. You can do this with anything. Just have your agent complete a task the way you want it, then tell it to save that as a new skill so you can use it again whenever you need to. So, I'm going to go ahead and send this. So, right here you can see it's actively using these skills, and we'll wait for a few minutes for it to come back with the All right, and here is the result. It came back with the top three AI tools that would make for an interesting tutorial video, then it gives me its best overall recommendation, and it also lets me know that it created a new skill for this exact workflow. By using this exact method right here, you can create new skills specifically for your agent, so that way you're not starting from scratch every single time. Now, let's make sure that this new skill actually was saved properly. So, I'll say, "Show me the entire list of skills again." Then I'll scroll up right here, and then under media, you see this new skill called YouTube video research, and that shows us that our new skill was saved properly. So, now the next time that I ask my Hermes agent for something similar, it's going to pull up this skill and use it automatically. So, think about what just happened. Hermes didn't just do the task, it saved how it did it, and now that skill is there permanently. Next time I ask for something similar, it already knows exactly what I want done. You can do this with anything. Research, trip planning, weekly reports, whatever it is, the pattern is the same. So, skills are how Hermes learns how to do things, but there's another important piece. How does it remember things about you specifically? Most AI tools forget you the moment you close your tab. You spend 20 minutes giving it context, your preferences, and projects, how you like things done, and then it's all gone. Hermes, however, can fix that. >> [music] >> To access the built-in memory and see what your agent remembers about you so far, that's stuff like your name, preferences, and anything else that you've shared, just ask it. I'll go ahead and type, "Show me what you remember about me so far." I'll send this. So, my Hermes agent doesn't know much about me just yet. However, here's what it does know. It knows that my name is Caleb, that I'm using Hermes through Telegram, I'm interested in Hermes skills, especially for video research and planning, and that I've asked for a repeatable workflow around researching trending AI tools. Here it tells me what it doesn't have yet, and it can also tell me about my current project interests, what it has saved as persistent memory, or it can save any new preference that you want it to remember going forward. So, here next I'm going to say, "Remember that I like my responses to be short and direct. No long explanations unless they are needed." So, I'll send this message to save a new preference into its memory. And then right here, you can see that Hermes has added it into its memory. To confirm that, I'm going to go ahead and pull up the memory again. So, I'll say, "Show me your memory again." And then right here, you can see that it added exactly what I said. It says that I prefer responses to be short and direct with no long explanations unless needed. Now, every response going forward with my agent will follow that preference. So, with this, you have a lot of control over what Hermes remembers about you. If you find it consistently doing something you don't like, you can pull up its memory, see if it saved that as a preference, and tell it to forget it. So, not only can you add new things into its memory, you can also remove them, too. And you can do all of this right here within Telegram. With Hermes, you're always the one deciding what it keeps in its memory. So, that's the built-in memory, and it works great right out of the box. But, there's also an optional upgrade called Honcho. Here's the difference in plain terms. Built-in memory stores facts. Honcho actually thinks about you between conversations. It picks up on patterns, builds a model of how you work, [music] and learns what you corrected on. So, my personal recommendation is upgrading to Honcho if you want to bring your agent's memory to the next level. All right, so to configure the memory of your Hermes agent, let's switch back over to the Hostinger web terminal. Now, if you closed your terminal since the last step or you're coming back to this anytime later in the future, you'll need to get back into the Hermes container first. You need to do this anytime you want to run Hermes commands from within the terminal. If you're on the standard root user or starting a new session, Hermes commands won't work without it. And once again, you'll also find all of these different Hermes commands listed in the video description within the Hermes setup guide. So, here let's go ahead and run CD /docker /hermes-agent. And for this again, you'll have to go back over to your Docker manager, find these last four characters right here. I'll go ahead and copy them, switch back over to the web terminal, then type dash and paste those in right here. Press enter, then go ahead and run docker compose exec -it hermes-agent /bin /bash. Press enter, and now you're back inside the container. From here, to see your memory options, go ahead and run Hermes memory setup. Run this. And then right here, you can see all the different memory providers that Hermes supports. And right here, you can see that we're currently using the built-in memory. Up above here, you'll see Honcho, which is my personal recommendation. If you want to make the switch from built-in memory, just select Honcho and follow the steps to create your account, set up billing, and connect your API key. I'm not going to walk through that exact process in this video, but it's straightforward, and you can go through it on your own from here. So, now that we have skills and memory set up, let's talk about automated jobs. This is where Hermes really starts to feel like an assistant and not just a chatbot. So, you're able to tell it to do stuff on a schedule without you having to prompt it first. This could be a morning briefing before you start work, a weekly recap at the end of a Friday, or just a reminder to stand up and stretch every couple hours. Anything that you'd normally have to remember to do yourself, you can have Hermes do. So, back here within Telegram, I'm going to ask my Hermes agent to set up a job for me. Here, I'll say, "Can you set up a weekly job that runs every Monday at 9:00 a.m. that researches the top three trending AI tools using the YouTube video research skill we created earlier?" And here, you'll notice that I'm not just creating a recurring scheduled job. I'm also combining it with the custom skill that we had Hermes create earlier. So, now we have a scheduled job running exactly how we want it to, powered by a custom skill, and backed by memory that keeps improving over time. Once we send this, our agent will start running autonomously, which I think is super cool. Let's go ahead and run this right here. And there we go. It's now been set up. We see the name, the schedule, the skill being used, where it's going to be delivered, and a job ID. It's also going to tell us when the job will be run next. And once you have a bunch of different jobs set up, it's also helpful to be able to see all of your jobs from a bird's-eye view. So, right here, I'll say, "Show me a list of all the current jobs." Send that. Right here, we can see that we have one job currently running. So, once you have a bunch of different jobs set up, you can quickly see them right here. And from here, if I wanted to add more jobs, pause this one, or remove it, I can manage all of that from right here. So, that's kind of the whole idea with your Hermes agent. You install or create new skills, set up automated jobs, and let your Hermes agent start handling tasks for you automatically. You can start small like I did here with my AI tools research, and then keep scaling from there based on what you need your agent to do. Here's one more important thing. Let's quickly talk about money. Hermes itself is free, and the VPS is very affordable. But, the AI model that you use charges you per message, and it can add up quickly if you're not paying attention. Here's how to think about it. You've got expensive models that are powerful for complex thinking, mid-tier models for most daily tasks, and cheap models that are perfect for routine stuff for scheduled jobs. The [music] rule of thumb here is pretty simple. Don't just run everything on the most expensive model. Use the expensive one for hard work, and the cheaper one for everything else. And if you remember when we set up Open Router at the start, I had to set a $50 spending limit. That right there is your safety net. No matter what happens, you can't go over that number. Now, from here to switch our model, I'll go ahead and type {slash} model, and go ahead and send it. Right here, we see our current model and the provider. Right here, we also see Open Router. So, let's go ahead and click on this. And here, I can easily select a new model. If I wanted to switch over to Opus, I could do that here, and I can also click on next to see the rest of the list. So, with this, it makes it super easy to change out your models. So, now what if I want specific jobs to utilize different models? My default model is GPT-5.4-mini. But, what if I want my weekly AI tools research job to run on a different model? Maybe something a little bit more powerful. That's something that you can also do with your Hermes agent. So, right here, I'm going to say, "I want the weekly AI tools research job to run on a different model. Change it to run on Opus 4.6." Send that. All right. So, now only the weekly AI tools research job is running on Opus 4.6, while my default model for regular chatting with my Hermes agent is still GPT-5.4-mini. This way, I can get the extra thinking power where I need it without driving up the cost of my everyday interactions. You can do this with any job as well. If it needs more horsepower, switch it to a more powerful model. If it's something simple, you can drop it down to a lighter model. It all depends on the scope of the job that you're creating. Now, before I finish up here, I want to quickly go over four more things that Hermes can do. We're not going to set these up step-by-step in this video, but I just want you to know that they exist because they can be pretty cool. The first thing here is voice messages. Hermes is able to transcribe voice memos that you send on Telegram, and it can even talk back using text-to-speech. Voice transcription is able to work right away. For spoken replies, just ask Hermes to enable text-to-speech. The default option is free, and it takes less than a minute to set up. The second thing is other messaging platforms. We set up Telegram because it's the easiest, but Hermes also works with Discord, Slack, WhatsApp, Signal, and a bunch of others. Just run a Hermes gateway setup in the terminal, and it's going to give you a menu. It's the same agent with the same memory, same skills. It's just another way to reach you. Now, the third thing is the skills hub. Beyond the pre-installed skills, there's a whole library of community skills that you're able to browse. Just ask your Hermes agent to look up the specific skill you're wanting to install, see if it's available, and if it is available, go ahead and ask your Hermes agent to check it for safety, and then just ask what the next steps are to install it. And the fourth thing is MCP support. MCP is a way to plug outside tools directly into Hermes, like adding the official GitHub MCP server so it has direct access. The command is Hermes MCP add, and it's going to walk you through exactly how to set it up. If that sounds advanced, don't worry about setting this up on day one. Just know that it's there whenever you're ready. All right. Now, from here, let me give you a quick overview of the Hostinger dashboard. That way, you can see exactly where your VPS settings are. So, let's switch back over to Hostinger. And right here, we're inside of our Docker manager. If I click on manage right here, and then I scroll down, you'll be able to find your environment variables here. Click on this. This is where you'll find your admin username and password, and you can change out those values here as well. Then, if I go to the left here and click on overview, you'll see all the information regarding your actual VPS right here. You'll be able to see your resources listed here, along with some other guides. Then, on the left-hand side, you have your settings, your operating system and control panel, and then the most important tab right here, which is your backups and monitoring. Click into it right here. And here, you can see your backup schedule. It's currently set to weekly. And then below this, you can create a snapshot. I would recommend doing this anytime you're about to make a major change to your VPS. That way, if something breaks, you can quickly restore back to a working version. Then, on the left here, you also have your security and your DNS settings. And if you need help with any other problems, or you just can't find something within your dashboard, go to the top right and click on Ask AI. This is going to bring up the Cody AI, and it's honestly pretty helpful for answering basic questions. And it can also forward you to Hostinger customer support, just in case something has gone wrong. And with that, you now have your own Hermes agent up and running, learning from every task you give it, and ready to handle stuff for you on a schedule. And if you haven't set up your VPS yet, go to this link on screen right here, or click the first link in the video description for the lowest pricing available. Thanks for watching. If this video helped you out, leave a like and subscribe for future tutorials. And if you run into any problems, just drop a comment down below, and I'll do my best to help you out. With that said, I'll see you in the next video.