Hermes-agent-guide

This guide will help anyone set up a Hermes agent on Nous research.

Hermes Agent is an autonomous agent that lives on your server, remembers what it learns, and gets more capable the longer it runs.

Requirements

Linux / macOS / WSL

🪟 Native Windows Setup (No WSL Required)

Prefer to run Hermes directly on Windows without WSL? Follow the native setup guide here: → Hermes Agent — Native Windows Setup (PowerShell 7)

Steps

Step 1

WSL Setup ( THIS IS FOR WINDOWS SUBSYSTEM ONLY )

Head to WSL on Windows, open the terminal

wsl

Step 2 ( MAC AND LINUX )

Installing Dependencies

Clone the repo to install dependencies

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

Step 3

Setting up your Hermes agent

• Select ENTER to proceed

• Select Provider: this will enable you to connect to your chat model

• Select Open router and click on Open router

• Go to Openrouterai to get an API key

• Sign up with GitHub on Openrouterai

• Click on API keys

• Click on Create API key

• Enter a name

• Copy your API key

  

Step 4

Model Setting

• Select model

• Choose nvidia/nemotron-3-super-120b-a12b:free

• Click on Enter

  

  Step 5

  TTS settings

  • Select TTS provider

  • Keep Current

  • Select Terminal backend

  • Choose Local

    

Step 6

Agent Settings

• Click on Enter on Max Iterations ( 60 )

• Press ENTER to use the default tool

• Press ENTER for context compression

  

Step 7

Session Reset

• Select Inactivity plus Daily Reset
• Press ENTER for inactivity timeout
• Press ENTER for the daily reset hour

Step 8

Messaging Platforms

• Select Y to use the Telegram bot and press enter
• Create a bot via @BotFather on Telegram
• Click on Start
• Enter a name for your bot. I used HERMES CHAT BOT
• Now let's choose a username for your bot. It must end in bot
• Done, your bot will be created
• Let's link our bot using the bot token
• Copy the bot token and paste it into your terminal and press ENTER. Do not worry, it will not display the token
• Restrict who can use your token
• Go to @userinfobot on Telegram
• It will reply with your numeric ID (e.g., 123456789)
• Copy it and paste it into your terminal
• Click on Y to use ID for home channel and press ENTER
• Click on N to set up the Discord Bot
• Click on N to set up Stack Bot
• Click on N to set up Matrix Bot
• Click on N to set up MatterMost Bot
• Click on N to set up WhatsApp
• click on N to set up GitHub, GitLab, etc
• Done

NT: Why I said N for the rest of the socials is cause they are less important, telegram is the most important of them all

Discord Bot Setup

• Go to https://discord.com/developers/applications → New Application → give it a name

• Click Bot on the left → Reset Token → copy your bot token

• Scroll to Privileged Gateway Intents → toggle ON Message Content Intent → Save Changes

• Go to Installation → Guild Install → pick scopes bot and applications.commands → copy the invite link

• Open the link, invite the bot to your server

• In Discord: Settings → Advanced → Developer Mode ON → right-click your name → Copy User ID

• Run hermes gateway setup → select Discord → paste your token and user ID when asked

• Done. Tag the bot in any channel to start chatting.

• 

Step 9

Install the gateway as a systemd service? (runs in background, starts on boot)

• Click on Y and press ENTER to use the default
• Click on Y to start the service

Step 10

SYSTEM MD SERVICES FOR HERMES

• You will see TOOLS for CLI, screenshot it and keep it for future references
• Press ESC to cancel
• Choose a provider, select ENTER to use the default
• Select FAL AI for image generation, head to FALAI
• Click on sign up with github
• Click on for myself and create with AI tools, and click on continue
• Click on Add Key
• Enter the relevant details
• Create key
• Copy your key and save it
• Paste it into your terminal and press ENTER
• Choose a voice provider, select the default to use your device's TTS
• Select Search Provider, select FIRECRAWL
• Sign up on FIRECRAWL with GitHub, use a VPN if the site does not respond
• Complete the onboarding tasks
• Choose the free plan
• Click on API keys
• Click on Create
• Enter the details
• Copy the key and paste it into your terminal and press ENTER
• Done

SYSTEM MD SERVICES IS FULLY CONFIGURED

Step 11

Interacting with HERMES

• You will see Would you like to install the gateway as a background service?
• Click on Y
• HERMES INSTALLATION IS COMPLETE AND IS LIVE ON TELEGRAM
• Reload your shell to use the hermes command:

Step 12

Interacting with HEMRES PART 2

• Use Hermes to start chatting with your agent
• Enter any input to start your message
• Done, your Hermes agent is successfully installed, and you can start chatting and it also works on your desktop background

Optional: Run Hermes on a VPS (24/7)

Want Hermes online 24/7 even when your PC is off? Put it on a cheap VPS.

This guide uses Tencent Cloud Lighthouse, it's affordable and beginner-friendly.

• Go to https://console.tencentcloud.com/lighthouse → click Sign Up Free to create an account
• After logging in, click Buy and select Lighthouse
• Pick a region close to you and choose Ubuntu as the OS
• Choose a plan, the Linux Starter (2 vCPU, 2 GB RAM) works perfectly and is very cheap
• Complete payment and wait for the instance to deploy
• Once ready, click Log In on your instance to open the web terminal, or SSH in
• Run the install:

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

• Go through the setup wizard, same as on your local machine
• Run hermes gateway setup to connect Telegram or Discord
• Start it in the background:

  hermes gateway --daemon
  

• Done. Your agent is live 24/7, chat with it anytime from your phone.

If you have any questions, dm MYSTIQUEMIDE

MADE WITH ❤️❤️ BY MYSTIQUEMIDE