Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.lmsa.app/llms.txt

Use this file to discover all available pages before exploring further.

LM Studio is a desktop application for Windows and macOS that lets you download and run large language models locally, then expose them over a local HTTP server. LMSA connects to that server over your Wi-Fi network, so your conversations never leave your home or office.

​
Before you start

Both your PC and your Android device must be connected to the same Wi-Fi network. LMSA communicates with LM Studio over your local network, so this is required for the connection to work.

​
Set up LM Studio on your PC

1

Install LM Studio

Download and install LM Studio from lmstudio.ai on your Windows or macOS computer.
2

Load a model

Open LM Studio, go to the Discover tab, and download a GGUF model (for example, Llama 3 or Mistral). Once downloaded, load it in the model selector at the top of the window.
3

Start the local server

Click the Developer tab (the <-> icon in the left sidebar) and press Start Server. The server runs on port 1234 by default. You will see a green indicator and a URL like http://localhost:1234.
4

Find your PC's local IP address

You need the PC’s IP address on your local network — not localhost or 127.0.0.1.
  • Windows: Open a terminal and run ipconfig. Look for the IPv4 Address under your Wi-Fi adapter (for example, 192.168.1.45).
  • macOS: Open System Settings → Wi-Fi → Details and copy the IP Address field.

​
Connect LMSA to LM Studio

1

Open Settings in LMSA

Tap the settings icon in the top-right corner of the LMSA home screen.
2

Select Local Server

On the Connection step, tap Local Server. Make sure the Ollama toggle is turned off — LM Studio uses a different API path.
3

Enter your PC's IP and port

Tap Configure, then enter your PC’s local IP address (for example, 192.168.1.45) and port 1234. Tap Save.
4

Start chatting

LMSA will detect the loaded model automatically. Return to the chat screen and start a conversation.
LMSA automatically detects which API version your LM Studio installation supports — both the newer /api/v1/ path (LM Studio 0.3.6 and later) and the legacy /v1/ path are handled transparently.

​
Optional: LM Studio API token

If you have enabled authentication in LM Studio’s server settings, you will need to provide an API token in LMSA. In LMSA Settings, on the Local Server panel, tap Token next to the key icon and paste your LM Studio API token. Leave this field empty if your server does not require authentication.

​
MCP servers

If you use LM Studio’s MCP (Model Context Protocol) server integrations, you can configure them in LMSA as well. On the Local Server panel, tap MCP to add your integrations. See MCP integrations for full details.

​
Troubleshooting

  • Confirm LM Studio’s local server is running (the green status indicator is visible in the Developer tab).
  • Double-check that you entered the PC’s local network IP (for example, 192.168.1.x), not localhost.
  • Verify both devices are on the same Wi-Fi network. Mobile data or a guest network will not work.
  • Check that your PC’s firewall is not blocking port 1234.
Make sure a model is loaded in LM Studio before connecting. LMSA reads the list of available models from the running server; if no model is loaded, the list will be empty.
Response speed depends on your PC’s hardware. Larger models require more RAM and CPU or GPU capacity. Try a smaller quantized model (for example, a Q4 variant) if responses are taking too long.
If you regularly switch between multiple PCs or network locations, use LMSA’s Saved Presets feature (in the Connection settings panel) to save and switch between server configurations without retyping your IP and port each time.