
Odysseus
ai
From $20 / month ยท billed monthly
Odysseus brings chat, documents, notes, tasks, and a calendar into one browser workspace. Connect an external model API using your own provider credentials for AI chat and research. Model API usage is billed separately by your provider. Local model installation, shell execution, CLI integrations, and MCP servers are disabled on this hosted edition. A fixed, single-threaded MiniLM embedding model supports document and memory search. The hosted policy source is available at /zenith-hosted-policy.py on your instance.
- ai
- chat
- documents
- notes
- tasks
- research
Screenshots
Frequently asked
- How much does it cost to host Odysseus?
- Hosting Odysseus on Zenith is $20 / month, billed monthly until you cancel.
- How long does Odysseus take to deploy?
- Most apps are live on your own subdomain in about two minutes after you deploy.
- Do I need to manage a server to run Odysseus?
- No. Zenith runs Odysseus for you, with no servers to provision or maintain.
- Can I use my own domain with Odysseus?
- Yes. Each deployment gets a Zenith subdomain, and you can map your own custom domain to it afterwards.
- Can I manage Odysseus with Claude or another AI agent?
- Yes. Zenith has a Model Context Protocol (MCP) server at https://zenith.hosting/mcp. Add the endpoint from Claude, Cursor or another MCP client, connect your account, and ask the agent to deploy, configure, inspect or delete your Odysseus deployment. The agent can read and set environment variables. Write-only secrets stay hidden; generated credentials may be returned. Deployment inspection reports live status, HTTPS reachability and summarized resource usage without exposing credentials. A paid deployment requires price approval. Deletion permanently removes the deployment and its data after separate confirmation. Setup is at https://zenith.hosting/connect.
Related apps
AFFiNE
Bring documents, whiteboards, and databases into one workspace. Move from a rough idea to an organized plan without switching tools.
- docs
- notes
- whiteboard
- +4
Bifrost
Put one OpenAI-compatible API in front of your model providers. Route requests, manage keys, and track usage from one console.
- gateway
- proxy
- openai
- +5
BookStack
Turn scattered team knowledge into a wiki with shelves, books, chapters, and pages. Write in a visual editor and search across everything you have written.
- wiki
- documentation
- knowledge-base
- +2


