Ollamarama is a powerful AI chatbot for the Matrix chat protocol powered by Ollama. Transform your Matrix channels with an AI that can roleplay as virtually anything you can imagine.

• 🎭 Dynamic Personalities: Switch between different AI personalities on the fly
• 👥 Per-User Chat History: Each user maintains their own conversation context
• 🔒 Channel Isolation: Conversations are separated by channel and user
• 🤝 Collaborative Mode: Users can interact with each other's chat histories
• 🛠️ Admin Controls: Model switching and global reset capabilities
• 🎯 Custom Prompts: Use your own system prompts for specialized interactions

• 💬 IRC Version: ollamarama-irc
• 🖥️ Terminal Version: ollamarama

Install and familiarize yourself with Ollama to run local LLMs.

curl https://ollama.com/install.sh | sh

Download the models you want to use from the Ollama library.

🎯 Recommended model:

ollama pull qwen3

pip install -r requirements.txt

This installs matrix-nio with encryption support so the bot can work in encrypted rooms.

1. Set up a Matrix account for your bot
2. Get your server URL, username, and password
3. Update the config.json file with your credentials:

{
    "matrix": {
        "server": "https://matrix.org",
        "username": "@your_bot:matrix.org",
        "password": "your_password",
        "channels": ["#your-channel:matrix.org"],
        "store_path": "store"
    }
}

python ollamarama.py

• Personality Examples: Try detective, pirate, shakespeare, helpful assistant, sarcastic critic
• Collaborative Mode: Use .x username to continue someone else's conversation
• Custom Prompts: Perfect for specialized tasks like code review or creative writing

• This bot supports end-to-end encryption (E2E) in Matrix rooms using matrix-nio[e2e] and a built-in device verification system.
• You must have libolm installed and available to Python for E2E to work.
• On Windows, you need to build and install libolm from source for encryption support. If you do not need encrypted rooms or have issues with libolm, use the files in the no-e2e/ folder, or run it using Windows Subsystem for linux (WSL).

This project is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0). See the LICENSE file for details.

Contributions are welcome! Feel free to:

• 🐛 Report bugs
• 💡 Suggest features
• 🔧 Submit pull requests

If you find this project useful, please consider giving it a star on GitHub!