Hermes Agent VPS Setup Guide for Beginners

About 8 min

Hermes Agent VPS Setup Guide for Beginners

Hermes Agent is most useful when it is always available. You can test it on a laptop, but a local setup stops when your computer sleeps, disconnects when your home network changes, and becomes awkward when you want scheduled tasks, messaging integrations, or long-running automation.

That is why many users move Hermes Agent to a VPS. A VPS gives you a small cloud server with a public IP address, root access, stable networking, and 24/7 uptime. In this guide, I will walk through the beginner setup process: choosing a server, connecting with SSH, preparing Linux, running Hermes Agent, keeping it online, and handling basic security.

What You Need Before Starting

You do not need to be a Linux expert, but you should be comfortable copying commands into a terminal.

Before you begin, prepare:

A VPS running Ubuntu 22.04 or Ubuntu 24.04
SSH access to the VPS
A domain name, optional but recommended for browser access
An API key for your LLM provider, such as OpenAI, OpenRouter, Anthropic, or another supported provider
Basic terminal access from your computer

For a first Hermes Agent server, I recommend:

Usage	CPU	RAM	Storage	Notes
Testing	1 vCPU	2GB	40-50GB SSD	Good for learning and light workflows
Daily use	2 vCPU	4GB	50GB+ NVMe	Better for multiple tools and integrations
Heavier workflows	4 vCPU	8GB+	80GB+ NVMe	Better for multiple agents or frequent jobs

Hermes Agent usually does not require a GPU if you use external LLM APIs. The VPS mainly runs the agent, memory, tools, integrations, and background services.

Recommended Beginner Option: LightNode Hermes Agent VPS

If you want the fastest beginner path, I recommend LightNode Hermes Agent VPS.

The main reason is simple: it is a pre-configured Hermes Agent VPS. Instead of spending your first hour installing dependencies, fixing version conflicts, and checking service logs, you can deploy a ready-to-run environment and start testing the agent quickly.

LightNode's Hermes Agent VPS page lists:

Hermes Agent pre-installed and ready to run
Deployment usually within 1-2 minutes
Hourly billing, so you can test without committing to a full month
40+ global locations for lower latency
Dedicated IP for API calls, webhooks, and integrations
NVMe SSD storage

The entry plan is practical for beginners:

Plan	CPU	RAM	Storage	Bandwidth	Monthly equivalent
Start	1 vCPU	2GB DDR4	50GB NVMe SSD	1TB	$7.71/month
Agency	2 vCPU	4GB DDR4	50GB NVMe SSD	2TB	$14.70/month
Premium	4 vCPU	8GB DDR4	50GB NVMe SSD	3TB	$27.70/month
Enterprise	8 vCPU	16GB DDR4	50GB NVMe SSD	4TB	$52.70/month

For a beginner, start with the 2GB plan if you only want to learn Hermes Agent. Choose 4GB if you plan to connect multiple tools, messaging channels, scheduled tasks, or several workflows.

👉 Deploy LightNode Hermes Agent VPS

Step 1: Create Your VPS

If you use a normal VPS provider, create a new Linux server with:

Ubuntu 22.04 LTS or Ubuntu 24.04 LTS
A data center near you or your users
SSH key login if available
At least 2GB RAM

If you use LightNode's Hermes Agent VPS, choose the Hermes Agent image or landing page deployment option, select your region and plan, then create the server. After deployment, copy the server IP address from the dashboard.

Step 2: Connect to the VPS with SSH

On macOS or Linux, open Terminal. On Windows, use PowerShell, Windows Terminal, or an SSH client.

Connect as root:

ssh root@YOUR_SERVER_IP

Replace YOUR_SERVER_IP with the IP address from your VPS dashboard.

If your provider gives you a password, SSH will ask for it. If you configured an SSH key, it should log in directly.

After login, check the system:

whoami
lsb_release -a
free -h
df -h

You should see the Linux version, available memory, and disk space.

Step 3: Update the Server

Always update a new VPS before installing anything:

sudo apt update
sudo apt upgrade -y

Install basic tools:

sudo apt install -y curl wget git unzip ca-certificates gnupg ufw htop

These tools help with downloads, Git repositories, firewall setup, and basic monitoring.

Step 4: Create a Non-Root User

For a real server, avoid doing everything as root.

Create a new user:

adduser hermes
usermod -aG sudo hermes

Switch to the new user:

su - hermes

From now on, run administrative commands with sudo.

Step 5: Set Up a Basic Firewall

Allow SSH first, then enable the firewall:

sudo ufw allow OpenSSH
sudo ufw enable
sudo ufw status

If Hermes Agent exposes a web dashboard or API on a specific port, allow only the port you actually use. For example:

sudo ufw allow 3000/tcp

Do not open random ports "just in case." Keep the server small and predictable.

Step 6: Run the Hermes Agent Installer

The official Hermes Agent VPS guide uses an installer script. The installer handles the core runtime setup, including the required Node.js and Python pieces.

Run it as your non-root user:

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

After installation, reload your shell if the installer asks you to:

source ~/.bashrc

Or log out and SSH back in:

exit
ssh hermes@YOUR_SERVER_IP

Check that the hermes command is available:

hermes --help

If the command is not found, check the installer's final output. It may tell you which profile file to reload or which binary path to add.

Step 7: Configure Your Model Provider

Hermes Agent needs an LLM provider. The official setup flow includes:

hermes model setup

Follow the prompts and add your provider details. Depending on your setup, this may ask for:

Provider name
Model name
API key
Base URL, if you use an OpenAI-compatible provider such as OpenRouter

Keep your API keys private. If you need to edit configuration manually, check:

nano ~/.hermes/config.yaml

Your Hermes data normally lives under:

~/.hermes/

If you are using LightNode Hermes Agent VPS, Hermes Agent should already be installed. In that case, your first job is usually not installation. Instead, check the welcome instructions from the LightNode dashboard or SSH session, add your API key, and start using the pre-installed runtime.

Step 8: Keep Hermes Agent Running

For a VPS, you want Hermes Agent to keep running after you close SSH and after the server reboots. The official gateway flow is:

hermes gateway install
hermes gateway start

If your Hermes version supports auto-start directly, enable it:

hermes gateway enable-autostart

Check the service:

systemctl status hermes-gateway

If your install uses PM2 instead of systemd, the beginner pattern is:

npm install -g pm2
pm2 start hermes --name hermes-agent
pm2 startup
pm2 save

Use only one process manager for the same Hermes instance. If hermes gateway install creates a systemd service successfully, you do not need PM2.

Step 9: Access Hermes Agent Safely

If Hermes Agent exposes a browser dashboard or local web service, you may be able to open:

http://YOUR_SERVER_IP:3000

For a private test, that is fine for a short time. For regular use, set up a domain and HTTPS.

A common production-friendly pattern is:

Point agent.yourdomain.com to your VPS IP
Put Nginx or Caddy in front of Hermes Agent
Enable HTTPS with Let's Encrypt
Keep the Hermes Agent app port private if possible

With Caddy, a simple reverse proxy can look like this:

agent.yourdomain.com {
    reverse_proxy 127.0.0.1:3000
}

This gives you HTTPS automatically when DNS is configured correctly.

Step 10: Add API Keys and Integrations

Hermes Agent becomes useful when it can connect to models and tools. Common settings include:

LLM API key
Default model name
Memory or database path
Telegram, Discord, Slack, or webhook tokens
Browser automation settings
Tool permissions

Keep secrets in .env files or your provider's secret manager. Do not paste API keys into public GitHub repositories, public docs, screenshots, or chat logs.

After editing Hermes configuration, restart the gateway:

hermes gateway restart

If your installation is managed by systemd, you can also use:

sudo systemctl restart hermes-gateway

Step 11: Basic Maintenance

Once Hermes Agent is running, maintenance matters more than the initial install.

Use these commands often:

htop
df -h
free -h
systemctl status hermes-gateway
journalctl -u hermes-gateway -n 100 --no-pager

Update the server:

sudo apt update
sudo apt upgrade -y

Update Hermes Agent:

hermes update
hermes gateway restart

Back up your important data directory:

tar -czf hermes-backup.tar.gz ~/.hermes

For serious use, schedule backups to another server or object storage. A VPS is reliable, but it is still your responsibility to protect agent memory, configs, and workflow files.

Common Beginner Mistakes

Choosing a VPS with too little RAM

Hermes Agent may start on a 1GB VPS, but it leaves little room for logs, browser tools, package updates, and extra integrations. Start with 2GB minimum. Use 4GB if you want a smoother daily setup.

Exposing the dashboard without HTTPS

Do not leave an admin dashboard open on plain HTTP for long-term use. Use HTTPS, strong passwords, firewall rules, or a private tunnel.

Forgetting about API costs

The VPS is only one part of the cost. If Hermes Agent uses external LLM APIs, model calls may cost more than the server if you run heavy workflows.

Running everything as root

Root is convenient, but it increases risk. Create a normal sudo user and use root only when necessary.

No backups

If your agent stores memory, files, prompts, or workflow state, back it up. A reinstall is easy; losing agent data is not.

When Should You Use a Pre-Installed Hermes Agent VPS?

Manual setup is useful if you want to learn Linux and control every part of the stack. But for beginners, a pre-installed VPS can save a lot of time.

Use LightNode Hermes Agent VPS if:

You want Hermes Agent running quickly
You do not want to debug Python, Node.js, system services, or dependency issues
You want hourly billing for experiments
You need a dedicated IP
You want to choose from many global locations
You are building a real 24/7 AI assistant or automation workflow

Use a manual VPS setup if:

You want to customize every dependency
You already know Linux server management
You are building a non-standard deployment
You want to inspect each service yourself

For most beginners, the best path is to start with a pre-installed Hermes Agent VPS, learn the runtime, then customize later when you understand your workflow.

FAQ

What is Hermes Agent VPS?

A Hermes Agent VPS is a cloud server used to run Hermes Agent continuously. It gives the agent a stable public IP, persistent storage, and 24/7 uptime.

Do I need a GPU for Hermes Agent?

Usually no. If you use external LLM APIs, the VPS only runs the agent and integrations. A CPU VPS with 2GB to 4GB RAM is enough for most beginner setups.

What is the best VPS size for beginners?

Start with 1 vCPU and 2GB RAM for testing. Choose 2 vCPU and 4GB RAM if you plan to use Hermes Agent daily or connect multiple tools.

Can I run Hermes Agent on a $5 VPS?

Yes, but expect limits. A very cheap VPS can work for simple testing, but 2GB RAM and fast SSD storage are a better baseline for a stable beginner setup.

Why use LightNode for Hermes Agent?

LightNode offers a pre-installed Hermes Agent VPS, hourly billing, NVMe SSD storage, 40+ global locations, dedicated IP, and fast deployment. It is a good fit if you want to skip manual setup and start testing quickly.

Is hourly billing useful for Hermes Agent?

Yes. Hourly billing is useful when you are testing agents, comparing model providers, or running temporary automation. You can stop the VPS when you are done instead of paying for unused time.

How do I keep Hermes Agent online after closing SSH?

Run it with the Hermes gateway service, systemd, or another process manager such as PM2. Do not rely on a foreground terminal session for long-running use.

Is it safe to expose Hermes Agent to the internet?

Only if you secure it properly. Use HTTPS, strong authentication, a firewall, and private access controls where possible. Never expose unauthenticated admin tools publicly.

How do I update Hermes Agent?

Use hermes update, then restart the gateway. Always back up ~/.hermes/ before major updates.

Can I run multiple Hermes agents on one VPS?

Yes, but each agent needs separate ports, data directories, and environment files. For multiple active agents, use at least 4GB RAM and monitor CPU, memory, and disk usage.

Final Thoughts

Running Hermes Agent on a VPS turns it from a local experiment into a real always-on assistant. The key is to keep the setup simple: use a stable Ubuntu server, start with enough RAM, secure SSH, protect your API keys, and run the agent with a persistent service method.

If your goal is to learn every detail, manual setup is worth doing. If your goal is to get Hermes Agent online quickly, LightNode Hermes Agent VPS is the cleaner beginner route because the environment is already prepared and you can focus on building workflows instead of fixing dependencies.