What’s available offline
When using offline mode with local servers, the following features are available:- Local LM Studio Models: Run models loaded in your LM Studio server on your computer
- Ollama Models: Use models from your Ollama installation
- Standard Chat: Send messages and receive responses from your local model
- File Attachments: Attach files to chats for analysis
- Text-to-Speech: Synthesize AI responses using available voices
- Templates & Character Cards: Use system prompts and roleplay characters
What’s not available offline
The following features require an internet connection and are not available when using local servers:- OpenRouter Cloud Models: Cannot access remote cloud AI services without internet
- Web Search API: Web search is not available in offline mode; the feature requires internet connectivity to query search engines
- Model Catalog Downloads: Cannot browse or download new models from online sources without an internet connection
Privacy benefits of offline mode
- Complete Privacy: Your conversations never leave your local network. No data is sent to cloud servers or third parties
- Full Control: You control all your data. Models run on hardware you own or control
- No Usage Tracking: Your chats are not monitored, logged, or analyzed by external services
- Sensitive Information Safe: Use LMSA offline for confidential documents, proprietary data, or personal information without privacy concerns
Enabling offline mode
1
Start your local server
Launch LM Studio on your computer and load a model, or start an Ollama server.
2
Connect in LMSA
Open LMSA Settings and configure your server’s local IP address and port.
3
Disconnect from internet (optional)
LMSA works whether internet is available or not, but for true offline operation, you may want to disable network connectivity on your mobile device.
4
Chat locally
Select your local model and start chatting with zero data leaving your network.