Self-hosted, open source AI assistant and persistent bot platform where you choose the model and sandbox, on web, desktop, iPhone and Android.
- Stars
- 94,028
- Forks
- 1,695
- Open issues
- 1,467
- Last commit
- 7 hours ago
What is it?
Odysseus is a self-hosted workspace that brings several everyday tools into one AI-driven interface. It covers chat with local or API-hosted models, agents with tool access (MCP, files, shell, skills and memory), multi-step web research that produces reports, and a writing-first document editor with AI edits.
The workspace also includes an IMAP/SMTP email client with triage, summaries and reply drafts, along with notes, tasks, reminders, scheduled agent tasks and CalDAV calendar sync. A Cookbook feature recommends models based on hardware and handles downloads and serving, and a Compare view tests models side by side.
It is deployed with Docker Compose using a published multi-arch image, and it includes an admin login with 2FA. The project's setup guide covers native installs, GPU notes, HTTPS and configuration. Auth is enabled by default for network-accessible deployments.
Key features
- 01Chat and agents with local or API models, tools, MCP and memory.
- 02Deep research that reads sources and generates reports.
- 03Side-by-side blind model comparison.
- 04Documents editor with AI edits, Markdown and CSV support.
- 05IMAP/SMTP email with triage, summaries and reply drafts.
- 06Notes, tasks, reminders and CalDAV calendar sync.
Use cases
- Running a private AI assistant on your own server instead of a hosted chat service.
- Managing email, notes and calendar in one self-hosted place alongside AI tools.
- Testing and comparing local models before choosing one for a workflow.
Quick start
- 1
Clone the repository
git clone https://github.com/odysseus-dev/odysseus.git - 2
Enter the project folder
cd odysseus - 3
Create your environment file from the example
cp .env.example .env - 4
Start the containers with Docker Compose, then open http://localhost:7011
docker compose up -d --build
Things to know
- Licensed under AGPL-3.0, a strong copyleft license that affects how you can distribute modified versions.
- The default dev branch gets the newest changes first, so it may be less stable than main.
- It is a self-hosted workspace with powerful local tools, so you must keep authentication enabled and avoid exposing service ports publicly.
Topics
FAQ
- What is odysseus?
- A self-hosted AI workspace combining chat, agents, deep research, documents, email, notes, calendar and local model workflows. Odysseus is a self-hosted workspace that brings several everyday tools into one AI-driven interface. It covers chat with local or API-hosted models, agents with tool access (MCP, files, shell, skills and memory), multi-step web research that produces reports, and a writing-first document editor with AI edits.
- Is odysseus free to use? What license does it have?
- odysseus is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.
- Is odysseus still maintained?
- Actively maintained. The last commit to odysseus was on Oct 20, 2026.
- What can I use odysseus for?
- Running a private AI assistant on your own server instead of a hosted chat service. Managing email, notes and calendar in one self-hosted place alongside AI tools. Testing and comparing local models before choosing one for a workflow.
- How do I install odysseus?
- Clone the repository: git clone https://github.com/odysseus-dev/odysseus.git Enter the project folder: cd odysseus Create your environment file from the example: cp .env.example .env Start the containers with Docker Compose, then open http://localhost:7011: docker compose up -d --build
- Where can I get odysseus?
- The source code and installation instructions are on GitHub: https://github.com/odysseus-dev/odysseus
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Self-hosted control center for running coding agents like OpenHands, Claude Code and Codex across local, remote and cloud backends.
Self-hosted personal AI agent framework in Python with a browser WebUI, terminal client, chat app channels, tools, memory and MCP support.
Open-source personal AI assistant and Agent Harness that plans tasks, runs tools and skills, and keeps memory across web and IM channels.
An agent-native, self-hostable AI learning workspace for personalized tutoring, Q&A, quizzes, research and visualization, built on Python and Next.js.
Self-hostable AI second brain that answers questions from the web and your own documents, with custom agents and scheduled automations.