Home Assistant Ollama Conversation Integration

A Home Assistant custom integration that provides conversation agent capabilities using Ollama with full tool/function calling support for device control.

✨ Features

• 🤖 Local AI Processing: Run language models locally via Ollama
• 🔧 Tool/Function Calling: Native support for Home Assistant device control
• 💡 Smart Home Control: Control lights, climate, and other devices through natural language
• 🔄 Conversation History: Maintains context across multiple turns
• ⚙️ Configurable Parameters: Adjust temperature, context window, and other model settings
• 🎯 Optimized for Function-Capable Models: Designed to work with tool-capable models like Home-FunctionGemma-270m

🚀 Quick Start

Prerequisites

1. Ollama Server: Running locally or on your network
2. Compatible Model: Download a tool-capable model:

   ollama pull home-functiongemma-270m

Installation via HACS (Recommended)

1. Open HACS in Home Assistant
2. Go to "Integrations"
3. Click the three dots menu → "Custom repositories"
4. Add this repository URL
5. Search for "Ollama Conversation" and install
6. Restart Home Assistant

Manual Installation

1. Copy the custom_components/ollama_conversation folder to your Home Assistant config/custom_components/ directory
2. Restart Home Assistant

Configuration

1. Go to Settings → Devices & Services
2. Click "+ Add Integration"
3. Search for "Ollama Conversation"
4. Enter your Ollama server URL (default: http://sanctuarymoon.local:11434)
5. Select your model from the dropdown
6. Configure optional parameters (temperature, context window, etc.)

📖 Documentation

See the detailed README for:

• Usage examples
• Supported device types
• Architecture details
• Troubleshooting guide
• API documentation

🆚 Comparison with Alternatives

vs. home-llm Integration

Feature	home-llm	ollama_conversation
Ollama Support	✅ Basic	✅ Full API
Tool Calling	❌ Limited	✅ Native
UI Configuration	❌ YAML only	✅ Config Flow
Model Switching	❌ Manual	✅ Dropdown
Error Handling	⚠️ Basic	✅ Comprehensive

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Add tests for new functionality
4. Submit a pull request

📝 License

MIT License - See LICENSE file for details

🙏 Credits

• Built by goose (Block's AI agent)
• Inspired by Home Assistant's bedrock integration
• Compatible with Home-FunctionGemma-270m model

🔒 Privacy

This integration processes everything locally. Your conversations never leave your network when using Ollama locally.

💬 Support

• 📖 Documentation
• 🐛 Issues
• 💬 Home Assistant Community