Hermes Agent on a VPS: Self-Host Hermes Anywhere

Why Self-Host Hermes?

Most Hermes users are best served by OpenClaw Launch — managed hosting with zero-ops deploys, automatic updates, and included AI credits. But self-hosting makes sense when:

• You need data to stay within a specific jurisdiction or on hardware you control
• You want to run Hermes on existing infrastructure you already pay for
• You need custom network routing, firewall rules, or VPN integration
• You want to run Hermes alongside a local Ollama or other private model server
• You have technical requirements that the managed platform does not cover

Recommended VPS Specs

These specs comfortably run the Hermes Docker image, the optional dashboard process, and have headroom for memory-backed vector search if you use Hermes's session memory feature. Providers like Hetzner, DigitalOcean, and Linode all offer 2 vCPU / 4 GB instances in the $6–$12/month range.

Step 1: Install Docker

Hermes Agent runs as a Docker container. Install Docker on your VPS:

# On Ubuntu 22.04 / 24.04
curl -fsSL https://get.docker.com | sh
sudo usermod -aG docker $USER
# Log out and back in for group changes to take effect

Verify Docker is working: docker run hello-world

Step 2: Pull the Hermes Image

The Hermes Agent image is hosted on GitHub Container Registry (GHCR). Always pin to a specific version tag rather than :latest to avoid unexpected behavior from auto-updates:

# Pull a specific pinned version (check the upstream repo for the
# current recommended tag — never use :latest in production)
docker pull ghcr.io/nousresearch/hermes-agent:YYYY.M.DD

# Verify the image is available
docker images | grep hermes-agent

Find the current recommended image tag in the upstream Hermes repository.

Step 3: Create a Persistent Data Volume

Hermes stores its configuration, session credentials, and memory data at /opt/data inside the container. Mount a host directory there to persist data across container restarts and updates:

# Create a directory on the host for Hermes data
mkdir -p /home/$USER/hermes-data

# The volume mount in your docker run command:
# -v /home/$USER/hermes-data:/opt/data

Without a persistent volume, Hermes configuration, WhatsApp/WeChat sessions, and memory data are lost every time the container is replaced.

Step 4: Run the Container

Start Hermes with your persistent data volume, the required environment variables, and the ports it needs:

docker run -d \
  --name hermes-agent \
  --restart unless-stopped \
  -v /home/$USER/hermes-data:/opt/data \
  -p 3000:3000 \   # Gateway port (Telegram/Discord/WhatsApp adapter)
  -p 9119:9119 \   # Hermes web dashboard
  -e OPENROUTER_API_KEY=sk-or-... \
  ghcr.io/nousresearch/hermes-agent:YYYY.M.DD \
  hermes gateway run

Replace YYYY.M.DD with the pinned version tag you pulled. The gateway runs on port 3000; the web dashboard runs on port 9119. The dashboard process may need to be started separately via docker exec — consult the upstream README for the exact startup procedure.

Step 5: Configure Provider and Model

After the container is running, set your inference provider and default model:

docker exec hermes-agent hermes inference set openrouter
docker exec hermes-agent hermes model set anthropic/claude-sonnet-4.6

Or edit /home/$USER/hermes-data/config.yaml directly and restart the container. See the model-specific guides for other provider options: Anthropic, DeepSeek, Grok, and others.

Firewall and Network Setup

For production self-hosted Hermes:

• Restrict port 9119 (dashboard) to your own IP with a firewall rule — this port should not be publicly accessible.
• Put a reverse proxy in front of port 3000 if you need HTTPS for your gateway domain. Caddy and nginx are both good options.
• Enable UFW or your provider's firewall to limit incoming traffic to only the ports you need.

Updating Hermes

To update Hermes to a new version, pull the new image tag and recreate the container. Your data volume persists:

docker pull ghcr.io/nousresearch/hermes-agent:NEW.VERSION.TAG
docker stop hermes-agent
docker rm hermes-agent
# Re-run the docker run command with the new tag

Self-Hosted vs OpenClaw Launch

For most individuals and small teams, OpenClaw Launch offers a significantly better experience at comparable or lower total cost. Self-hosting is the right choice when you have specific data residency, integration, or operational requirements that managed hosting cannot meet.

What's Next?

• Install Hermes Agent — Full installation walkthrough including Docker prerequisites
• Deploy Hermes Agent — Deployment patterns for different environments
• Hermes Agent + MCP — Extend self-hosted Hermes with MCP tool servers
• Hermes Agent + Anthropic — Configure Claude on your self-hosted instance