Ontheia
Ontheia is a self-hosted, open-source AI agent platform that lets you run specialized agents, tools, memory, and workflows on your own infrastructure. It solves the problems of fragmented AI setups and loss of data control by providing a governed, integrated environment where agents can collaborate, remember, and act autonomously. It is for developers and organizations that want sovereign AI with full control over data and rules.
✨ Key features
- Specialized agents and reusable Agent Skills
- Delegation and coordination between agents
- MCP-native tool integration
- Governed persistent memory with RAG and pgvector
- Visual Chain Engine for no-code workflows
- Built-in scheduling for recurring jobs
🎯 Use cases
- Build a personal assistant that manages tasks and remembers context
- Create automated workflows that delegate subtasks to specialized agents
- Connect agents to external tools via MCP for data retrieval and actions
- Run scheduled jobs that resume existing chats
- Deploy a multi-user AI platform with strict data governance
📦 Installation
🧰 Requirements: Linux or macOS (Windows via WSL2), Docker 24+, Docker Compose v2, 2 GB RAM (4 GB recommended), 5 GB disk (10 GB recommended), ports 8080 and 5173, openssl, curl, jq, and at least one AI provider API key (or local Ollama).
Guided setup (recommended):
curl -fsSL https://get.ontheia.ai | bash
Manual setup:
git clone https://github.com/Ontheia/ontheia.git
cd ontheia
cp .env.example .envdocker compose up -dThen visit http://localhost:5173.
🚀 Usage
After installation, open http://localhost:5173 in your browser. The system ships with two agents: the Ontheia Guide and a Personal Assistant. You can start chatting with them to explore capabilities.
⚠️ Good to know
Requires at least one AI provider API key to chat; memory features require an embedding-capable provider. The project is dual-licensed under AGPL-3.0 and a commercial license.
❓ FAQ
What are the system requirements?
Linux/macOS (Windows via WSL2), Docker 24+, Docker Compose v2, 2 GB RAM (4 GB recommended), 5 GB disk, ports 8080 and 5173, and openssl, curl, jq.
Do I need an API key?
Yes, at least one AI provider API key is required to chat, such as Anthropic, OpenAI, or a local Ollama instance. For memory/RAG features, you need an embedding-capable provider.
Can I use local models?
Yes, Ontheia supports Ollama and any OpenAI-compatible model, so you can run entirely on local models if desired.
Is Ontheia open source?
Yes, it is dual-licensed: AGPL-3.0 for open-source use and a commercial license for proprietary use.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.