Deploy and Host Hermes Agent

Hermes Agent is an open-source AI agent by Nous Research. This template deploys the official Hermes container image with an authenticated web dashboard, persistent state, and native OpenAI Codex/ChatGPT OAuth support.

About Hosting Hermes Agent

The template creates one Railway service from the public otopba/hermes-railway-template repository, exposes dashboard port 9119 over HTTPS, and mounts a persistent volume at Hermes' upstream data path /opt/data.

Dashboard username, password, and session signing are configured automatically. The password and signing secret are independently generated for every deployment. The optional OpenAI-compatible API server stays disabled by default.

After Railway reports SUCCESS, reveal HERMES_DASHBOARD_BASIC_AUTH_PASSWORD in the service Variables tab and sign in to the generated domain as admin.

To connect a ChatGPT subscription, open the service shell and run:

hermes auth add openai-codex

Complete the device-code sign-in shown by Hermes. Credentials are written to /opt/data and survive redeployments.

Why Deploy Hermes Agent on Railway

One-click deployment from the official Nous Research image
Persistent credentials, sessions, memory, skills, and configuration
HTTPS dashboard with generated authentication secrets
Healthchecked deploys and automatic restart on failure
Native Hermes OpenAI Codex OAuth flow, without storing a ChatGPT password
GitHub-backed source with pinned image tag and immutable digest

Common Use Cases

Run a personal always-on Hermes agent backed by a ChatGPT subscription
Use Hermes' web dashboard from multiple devices
Add Telegram, Discord, Slack, or other supported gateway integrations
Keep agent memory and OAuth credentials across container replacements
Test Hermes workflows without maintaining a VPS or Docker host

Dependencies for Hermes Agent Hosting

A Railway account with capacity for one service and one persistent volume
A ChatGPT account only if you choose the openai-codex provider
Optional credentials for messaging platforms or other model providers you enable later

Deployment Dependencies

Source: https://github.com/otopba/hermes-railway-template (main)
Base image: nousresearch/hermes-agent:v2026.8.3, pinned by digest
Public port: 9119
Healthcheck: /api/health
Persistent volume: /opt/data
Start command: /opt/hermes/docker/entrypoint-dispatch.sh gateway run

This is a community-maintained deployment template and is not affiliated with Nous Research, OpenAI, or Railway.