Deploy Odysseus
Self-hosted AI workspace with ChromaDB, SearXNG search and ntfy push
Just deployed
/data
Just deployed
/var/lib/ntfy
Just deployed
/app/data
Just deployed
Deploy and Host Odysseus on Railway
Odysseus is an open-source, self-hosted AI workspace. It brings chat with any LLM provider, agents with tools and MCP servers, deep research with web search, notes, document RAG, memory and reminders into one web app. You bring the models (OpenAI, Anthropic, OpenRouter, Ollama or any OpenAI-compatible API); Odysseus keeps your data on your server.
About Hosting Odysseus
Odysseus runs as a Python web app next to three helper services: ChromaDB stores the vectors for document search and memory, SearXNG gives agents and deep research private web search, and ntfy delivers reminders and notifications to your phone. This template deploys all four on Railway's private network, each from a Dockerfile that pins the official image by version and digest. Odysseus keeps its database, users, settings, uploads and logs on one volume, ChromaDB and ntfy each have their own, and SearXNG is stateless. Every password and token is generated at deploy time, and sign-in is always on.
Common Use Cases
- A private ChatGPT-style workspace for you or a small team, using your own API keys or self-hosted models
- Deep research and agent workflows that search the web through your own SearXNG instead of a paid search API
- Chat with your documents and notes (RAG), with reminders and agent results pushed to your phone through ntfy
Dependencies for Odysseus Hosting
- At least one LLM provider: an API key (OpenAI, Anthropic, OpenRouter and others) or a reachable Ollama or OpenAI-compatible endpoint, added in the app after you log in
- ChromaDB, SearXNG and ntfy, all included in this template
Deployment Dependencies
- Odysseus source and docs: https://github.com/odysseus-dev/odysseus and https://odysseus-dev.github.io/odysseus
- Template repo, changelog and update notes: https://github.com/aalfath/odysseus-railway-template
- ChromaDB: https://docs.trychroma.com
- SearXNG: https://docs.searxng.org
- ntfy: https://docs.ntfy.sh
Implementation Details
Services:
- odysseus (public): the web app on port 7000, health check
/api/health, volume/app/data. The admin account is created on first boot asadminwith the generatedODYSSEUS_ADMIN_PASSWORD; read it in the service's Variables tab. Changing that variable later doesn't change the password, use the account settings in the app. - chromadb (private): vector database, volume
/data. - searxng (private): metasearch engine with the JSON API Odysseus uses. No volume needed; its secret is generated.
- ntfy (public, login required): push notification server, volume
/var/lib/ntfy. Anonymous access is denied. Odysseus is pre-connected with a generated access token (Settings, Integrations). To get notifications on your phone, install the ntfy app, add your ntfy service's Railway URL as a server, and log in asadminwith the generatedNTFY_ADMIN_PASSWORDfrom the ntfy service's Variables tab.
After deploy: open the Odysseus URL, log in, and add your model provider under Settings. Optional features (email, Google OAuth, Brave or Tavily search, Hugging Face downloads) are set up in the app or with the variables listed in the upstream configuration reference.
Updates: the services build from the template repo. When it moves to a new Odysseus release, Railway shows an update notice in your project; read the repo's CHANGELOG.md first, then apply it. Your volumes are kept.
Why Deploy Odysseus on Railway?
Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.
By deploying Odysseus on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.
Template Content

