Installation Guide

Get Hermes Running

Hermes Agent v0.8.0 installs in about 60 seconds on Linux, macOS, or WSL2. One curl command, and the installer handles uv, Python 3.11, Node.js, ripgrep, and ffmpeg for you.

checklist System Requirements

javascript Node.js

Only git is required up front — Python 3.11, Node.js v22, uv, ripgrep, and ffmpeg are installed by the script

computer OS

Linux, macOS, WSL2 on Windows, or Android via Termux. Native Windows is not supported.

memory RAM

2 GB RAM minimum; more if you run local models

hard_drive Disk

~1 GB for the agent, its venv, Node modules, and built-in skills

extension Optional

Optional: Docker, SSH, Daytona, Singularity, or Modal for sandboxed command execution

wifi Network

Internet connection required for hosted model APIs. Local models via an OpenAI-compatible endpoint work offline.

The one-liner from the official docs. Recommended for most users on Linux, macOS, or WSL2. Install docs arrow_forward

WSL2 (Windows)

Native Windows is not supported. Run this inside a WSL2 shell after 'wsl --install'.

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

macOS / Linux / WSL2

Only git is required up front — the installer handles uv, Python 3.11, Node.js v22, ripgrep, and ffmpeg.

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

verified Verify Your Installation

Run the verification commands listed in the official docs.

1
$ hermes version

Prints the installed version (should be 0.8.0 or later).

2
$ hermes doctor

Runs a diagnostic check — Python, uv, dependencies, config files, and connectivity.

3
$ hermes status

Prints your current configuration: model provider, enabled tools, and connected gateways.

settings Initial Configuration

After installation, reload your shell, then run the three setup commands from the docs.

1

Choose a model provider

Run 'hermes model' to pick a provider. The README lists Nous Portal, OpenRouter, z.ai/GLM, Kimi/Moonshot, MiniMax, OpenAI, and custom endpoints. Anthropic and Mistral also ship as dependencies.

2

Enable the tools you want

Run 'hermes tools' to configure which of the 47 built-in tools are enabled. You can add or remove them later.

3

Run the full setup wizard

Run 'hermes setup' to walk through the complete wizard, including API keys and chat gateway connections (Telegram, Discord, Slack, WhatsApp, Signal, Email).

4

Send a test message

Try 'hermes chat -q "Hello! What tools do you have available?"' — straight from the official quickstart.

.env — Config lives in ~/.hermes/ — config.yaml for main settings, .env for API keys, plus subdirectories for sessions, logs, memories, skills, and cron jobs. See the env-vars reference at hermes-agent.nousresearch.com/docs/reference/environment-variables.

cloud VPS / Server Deployment

Hermes runs on any Linux VPS with 2 GB RAM and ~1 GB disk. The same one-liner works over SSH.

1

SSH into your server

Any Linux distro works — Ubuntu, Debian, CentOS, Arch, etc.

$ ssh user@your-server-ip
2

Run the installer

The script auto-detects your distro, installs Python 3.11 via uv, and sets up all dependencies.

$ curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
3

Enable a sandbox backend

On shared infrastructure, configure one of the sandbox backends (Docker, SSH, Daytona, Singularity, or Modal) so Hermes's shell commands run in isolation. See the README for backend-specific setup.

4

Set your API key and launch

Pick a model provider and start Hermes. Use tmux or screen for persistence — the process runs in the foreground.

$ hermes model && hermes

Tip — For a persistent daemon, run Hermes inside tmux/screen or write a systemd unit file. Check the official docs for daemon setup.

system_update_alt Upgrading Hermes

The README documents a built-in updater: 'hermes update'. There is no published PyPI package or Docker image, so other upgrade paths go through git. Upgrade guide arrow_forward

npm
$ hermes update
pnpm
$ cd ~/.hermes/hermes-agent && git pull && uv pip install -e ".[all]"
Git
$ curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
Docker
$ # No official Docker image is published yet

See github.com/NousResearch/hermes-agent/releases for what's in each version.

delete_forever Uninstalling Hermes

Hermes installs into ~/.hermes/hermes-agent/ and symlinks the binary into ~/.local/bin/hermes. Removing both takes two commands. Uninstall notes arrow_forward

npm
$ rm ~/.local/bin/hermes
pnpm
$ rm -rf ~/.hermes/hermes-agent
Git (Source)
$ rm -rf ~/.hermes/hermes-agent
Docker
$ # No official Docker image to remove

Remove user data (optional)

This deletes your sessions, memories, skills, cron jobs, and API keys. Back up anything you need first.

macOS/Linux $ rm -rf ~/.hermes
Windows PS> # Run inside your WSL shell — native Windows is not supported

build Troubleshooting

Taken from the install docs' troubleshooting table and common install script failures. More troubleshooting tips arrow_forward

Need Help?

Join the Nous Research Discord for installation support and community troubleshooting.

Stay in the Loop

Community updates on Hermes Agent releases, new skills, and integrations. No spam, unsubscribe anytime.