LM Studio setup

Do this on the computer that will run your model. LLMster uses the computer’s memory and processing power; installing the client on your phone does not move model inference onto the phone.

Install and choose a model

1. Download and install LM Studio for your host’s operating system. For a machine without a desktop environment, follow Headless setup.
2. Open LM Studio’s model discovery interface and download a chat model. Check its download size and memory requirements against your computer. Choose a model that explicitly supports vision or tools if you need those features.
3. Load the model in LM Studio and confirm it can answer a short test prompt. This separates host/model problems from the client connection.

LM Studio’s model download guide explains the available formats and downloads. A model’s maximum context and memory use also depend on its loading configuration.

Start the API server

Open Developer → Server Settings. Check the port; the usual starting value is 1234. Enable Serve on Local Network so your iPhone or iPad can connect, then start the server.

Copy the reachable URL shown by LM Studio. For example, http://192.168.1.20:1234 contains the host’s address and server port. Your actual address may differ.

Protect the connection

1. In Server Settings, enable Require Authentication.
2. Open Manage Tokens → Create Token. Name the token for your device and choose the permissions needed for chat and model management. Include integration access if you intend to use tools.
3. Copy the token while it is displayed. You will enter it in LLMster’s Auth Key field.

Network access

Authentication controls who can use the server; an HTTP address does not encrypt traffic. Use your trusted local network. This guide does not require forwarding a router port or exposing the server to the public internet.

See LM Studio’s authentication guide for token management. Authentication requires LM Studio 0.4.0 or newer.

Keep the host address stable

Prefer the address displayed by LM Studio. If you need to find it in your host settings:

macOS

Open System Settings → Network, select the active connection, and inspect its IP address.

Windows

Run ipconfig in a terminal and find the IPv4 Address of the active network adapter. Ignore disconnected adapters.

Linux

Open your network connection details and find the active interface’s IPv4 address. Use the interface on the same network as your Apple device.

A router’s DHCP reservation can keep that address from changing. If it changes later, update the saved URL in LLMster. Keep the host awake while chatting and allow the server port through its firewall.

Connect LLMster