Obsidian
Obsidian integration
Combine your notes with local, private AI by integrating Loc.ai with Obsidian.
What is Obsidian?
Obsidian is a writing app based on local Markdown files, linking, and a rich plugin ecosystem — often used as a "second brain".
Why Loc.ai + Obsidian?
- Privacy — Notes and LLM traffic stay on your devices when you use local serving
- Offline-friendly — Work without cloud AI dependencies once models are deployed
Prerequisites
- Loc.ai Link installed and device registered — registration guide
- Link running on the device
- Obsidian installed
Step 1: Upload a language model
Loc.ai supports GGUF language models — see Supported models.
- ML MODELS → Upload ML Model
- Upload a GGUF file or choose a shared language model
- Set Model Type to Language Models
Upload time depends on model size and bandwidth.
Step 2: Deploy
- In ML Models, Action on your model
- Select an online device with enough storage
- Deploy
Step 3: Serve
- Click Serve Model
- Choose a port (default
8100)
info
Use the resulting base URL (e.g. http://127.0.0.1:8100/v1) in plugins.
Step 4: Enable community plugins
Settings → Community plugins → turn on community plugins (exit Restricted Mode).
Back up your vault before enabling community plugins.
Use case: Smart Composer
Install: Smart Composer from community plugins.
Configure:
- Add custom provider → OpenAI Compatible
- Base URL:
http://<YOUR_DEVICE_IP>:<PORT>(per plugin docs) - Add models and set chat/apply models
tip
Multiple models on different ports → multiple providers.