---
title: "Deploy Hermes Agent"
description: "Deploy Hermes Agent: an open-source AI agent with dashboard and memory."
category: "AI/ML"
url: https://railway.com/deploy/hermes-agent-13
---

# Deploy Hermes Agent

Deploy Hermes Agent: an open-source AI agent with dashboard and memory.

**[Deploy Hermes Agent on Railway](https://railway.com/template/hermes-agent-13)**

Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/hermes-agent-13/manifest.json

- **Creator:** Andrea Taglia's Projects
- **Category:** AI/ML

## Template content

### hermes-agent

- **Source:** https://github.com/andreataglia/hermes-agent-railway-template
- **Start command:** `/usr/local/bin/hermes-railway-entrypoint gateway run`
- **Health check:** /api/status
- **Public domain:** Yes

## Documentation

# Deploy and Host Hermes Agent with Railway

Hermes Agent is an open-source AI agent from [Nous Research](https://nousresearch.com) that runs as a persistent service: a browser dashboard, long-term memory, skills, scheduled tasks, and messaging channels you can talk to from Telegram, Discord, Slack or WhatsApp. This template runs the official `nousresearch/hermes-agent` image with a volume attached, so the agent keeps its state across redeploys.

## About Hosting Hermes Agent

This template provisions a single Railway service from the official image, with s6 supervising the gateway so the agent restarts itself if a channel, tool or cron job crashes. A Railway volume is mounted at `/data`; Hermes stores everything it knows under `/data/.hermes` — config, sessions, memory, skills, profiles and channel credentials.

The dashboard is served on port `8080` behind a username and password, and the service exposes a healthcheck at `/api/status`. The image itself is stateless, so upgrading Hermes never touches your data.

## Why Deploy Hermes Agent

- **One click, no CLI.** The image, entrypoint, volume mount, port, healthcheck and restart policy are already configured.
- **State that survives redeploys.** Memory, sessions and channel credentials live on the volume, not in variables.
- **A dashboard, not a container you SSH into.** Configure models, channels, skills and cron jobs from the browser.
- **Self-hosted and portable.** Your data stays in your Railway project; the agent runs on open-source software you can move anywhere.
- **Upgrade on your schedule.** The Dockerfile pins a release by tag and digest; bump it and redeploy.

## Common Use Cases

- A personal assistant reachable from Telegram or WhatsApp that remembers context across conversations.
- Scheduled automations: morning briefings, inbox triage, monitoring and reports.
- A research or knowledge-base agent with persistent memory over your own material.
- A self-hosted agent sandbox for testing models, skills and tool integrations.

## Dependencies for Hermes Agent

- A model provider: a Nous Portal account, or an API key for OpenRouter, Anthropic, OpenAI or any OpenAI-compatible endpoint. Configure it in the dashboard after the first boot.
- Optional: a bot token for whichever messaging channel you want to connect.
- Railway resources: the default volume (5 GB) plus roughly 1 GB of RAM is a sensible starting point.

### Deployment Dependencies

| Variable | Required | Description |
|---|---|---|
| `ADMIN_USERNAME` | Yes | Username for the dashboard login. `admin` is a fine default. |
| `ADMIN_PASSWORD` | Yes | Password for the dashboard login. Choose something strong — this is the only thing protecting a publicly reachable dashboard. |

| Setting | Value |
|---|---|
| Volume mount | `/data` |
| Public port | `8080` |
| Healthcheck path | `/api/status` |
| Start command | `/usr/local/bin/hermes-railway-entrypoint gateway run` |

No model or messaging keys are needed to deploy — add them from the dashboard after the first boot.

## After deploying

1. Open the generated Railway domain and sign in with the credentials you supplied.
2. Connect a model provider in the dashboard. Nothing works until a model is configured.
3. Optionally connect a messaging channel and paste its bot token.
4. Anything you configure from the dashboard is written to the volume, so it survives the next redeploy and never appears in this template.

Do not detach or delete the volume: it holds the agent's entire memory and state.

## Upgrading

This template builds from [`andreataglia/hermes-agent-railway-template`](https://github.com/andreataglia/hermes-agent-railway-template), which pins a Hermes release by tag and digest in its `Dockerfile`. Bump that pin and redeploy to upgrade. Redeploying pulls the latest commit on `main`, so existing deployments can move to a newer release without recreating the template.

## Resources

- [Hermes Agent documentation](https://hermes-agent.nousresearch.com/docs)
- [Docker deployment guide](https://hermes-agent.nousresearch.com/docs/user-guide/docker)
- [Hermes Agent on GitHub](https://github.com/NousResearch/hermes-agent)

## Support

Issues with the agent or the image belong upstream in [NousResearch/hermes-agent](https://github.com/NousResearch/hermes-agent/issues). Issues with this Railway template (image pin, entrypoint, Railway settings) go to [andreataglia/hermes-agent-railway-template](https://github.com/andreataglia/hermes-agent-railway-template/issues).


## Similar templates

- [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform.
- [Weaviate](https://railway.com/deploy/weaviate-1) — Vector database with hybrid search and an API key-protected API
- [Möbius Personal AI OS](https://railway.com/deploy/mobius-v01-personal-ai-os) — Self-hosted AI agent that builds and runs apps for you, with memory.

Open this page in a browser: https://railway.com/deploy/hermes-agent-13
