Connect LLMster

First start your LM Studio server and have its address and authentication token ready.

Enter the connection details

1. Open LLMster. On first launch, follow setup; on later visits, open Settings → Server.
2. Enter Server URL. Use the base address and port, without an API endpoint such as /v1/chat/completions. If you omit the scheme, LLMster adds http://.
3. Enter Auth Key if authentication is enabled on the server. Paste the token itself, without adding Bearer.
4. Choose Test Connection. Allow local-network access when the system asks.

Choose the right address

Use your computer’s network address, such as http://192.168.1.20:1234. Your iPhone or iPad must be able to reach the computer on the network. localhost and 127.0.0.1 refer to the phone or tablet itself and will not reach the computer.

Substitute your actual port if it is different. 0.0.0.0 is a server listening address, not the host address to save in LLMster.

Understand the result

Connected means LLMster reached the server and retrieved its model catalog. The status includes the number of models found and loaded-model information. It does not guarantee that every model supports images, reasoning, or tools.

If no model is loaded, open a chat and choose a model. If the test fails, follow connection troubleshooting.

Authentication storage

The app saves the authentication token in the device Keychain. A library backup does not include it. Configure the server and token separately on another device.

Change servers

Update Server URL and Auth Key, then test again. Existing chat history stays on the device, but saved server-side continuation context may belong to the previous server. Follow any continuation prompt rather than assuming two servers share the same context.