If you use LM Studio as a local AI server, manually opening the application and starting the server every time Windows restarts quickly becomes inconvenient.

Fortunately, LM Studio can run its local LLM service automatically when you sign in to Windows. You can also disable this behavior if LM Studio keeps starting when you do not want it to.

This guide explains how to:

• Start LM Studio automatically on Windows

• Run the LM Studio LLM server on login

• Stop LM Studio from starting automatically

• Use Windows Startup settings as a fallback

• Run LM Studio headlessly in the background

• Automatically make models available to API clients

Can LM Studio Start Automatically on Windows?

Yes.

LM Studio includes a built-in option that can run its LLM server automatically when you log in to your computer. When enabled, LM Studio can continue running in the background rather than requiring you to manually open the interface each time. 

This is especially useful if you use LM Studio as a local API backend for applications such as:

• AI coding tools

• Local chat interfaces

• Automation scripts

• OpenAI-compatible applications

• Home servers

• Development environments

If your only goal is to have the LM Studio API available after you log in to Windows, the built-in option is usually the easiest solution.

Method 1: Enable LM Studio Server Auto Start on Windows

The recommended approach is to use LM Studio's own startup setting.

Step 1: Open LM Studio

Launch LM Studio normally on your Windows PC.

Step 2: Open Settings

Open the application settings.

You can also use:

Ctrl + ,

Step 3: Enable the LLM Server on Login

Look for the option that allows LM Studio to run the LLM server on login and enable it.

LM Studio's official documentation states that this setting starts the LLM service when you log in to the machine. With the option enabled, closing the main application window can leave LM Studio running in the system tray while the server continues operating in the background. 

Step 4: Start the Server Once

Open the Developer section in LM Studio and start the local server.

LM Studio normally provides a Start server control in the Developer tab. The server can also be started from the command line with:

lms server start

LM Studio saves the previous server state and can restore it when the application or service starts again.

After configuring these options, restart Windows and sign back in to test the setup.

Does LM Studio Need to Stay Open?

Not necessarily in the traditional sense.

When LM Studio's background service option is enabled, closing the main window can minimize LM Studio to the system tray instead of completely shutting down the LLM service. 

That means your local API can remain available while the main LM Studio interface stays out of the way.

This is useful if another application connects to LM Studio through an endpoint such as:

http://localhost:1234

You can therefore use LM Studio primarily as a background local AI service rather than as a desktop chat application.

How to Stop LM Studio From Starting Automatically on Windows

Some users have the opposite problem: LM Studio starts every time they log in, even though they no longer want it running in the background.

There are two places worth checking.

Option 1: Disable Auto Start Inside LM Studio

Open LM Studio Settings and disable the option that runs the LLM server on login.

This is the first setting to check because LM Studio itself can intentionally configure its background service to start when you log in. 

After disabling it, fully exit LM Studio and restart Windows to confirm that the server no longer launches automatically.

Option 2: Check Windows Startup Apps

Windows also maintains its own list of applications that can start automatically when you sign in.

Go to:

Start > Settings > Apps > Startup

Find LM Studio if it appears in the list and turn it Off.

Microsoft confirms that applications registered as startup apps can be enabled or disabled from this page. The same startup list is also available through Task Manager. 

You can alternatively open:

Task Manager > Startup apps

Then select LM Studio and choose Disable.

This is particularly useful if you disabled the LM Studio setting but the desktop application itself still launches when Windows starts.

Method 2: Add LM Studio to the Windows Startup Folder

If LM Studio does not appear in Windows Startup Apps, Windows provides another way to launch applications automatically.

Press:

Win + R

Then enter:

shell:startup

This opens the Startup folder for your Windows account.

You can place an application shortcut in this folder to have Windows launch it automatically when you sign in.

Microsoft documents shell:startup as the Startup folder for the current user. For startup behavior applying to all users, Windows also provides shell:common startup. 

However, for most LM Studio users, the application's built-in LLM server on login option is preferable if the real goal is simply to keep the local API server running.

The Startup folder method is more useful when you specifically want the LM Studio desktop application to open.

App Auto Start vs. Server Auto Start

This distinction is important.

Starting the LM Studio application and starting the LM Studio local server are not exactly the same thing.

If you simply add LM Studio to Windows Startup, Windows launches the application.

But if your real goal is to provide a local AI API for other software, what you actually need is the LLM server running.

LM Studio's server can expose local models through its own REST API as well as OpenAI-compatible and Anthropic-compatible endpoints. 

For that use case, configuring LM Studio as a background service is usually cleaner than forcing the full graphical application to open every time Windows starts.

Run LM Studio Headlessly on Windows

For users building a more permanent local AI server, LM Studio also offers a dedicated headless option called llmster.

LM Studio describes llmster as the standalone core of LM Studio designed to run without the desktop GUI. The project recommends it for server-style and headless deployments. 

On Windows, it can be installed from PowerShell using:

irm https://lmstudio.ai/install.ps1 | iex

Then the daemon can be started with:

lms daemon up

The local server can be started with:

lms server start

This approach makes more sense if your Windows computer functions primarily as an AI server and you rarely need to interact with LM Studio's desktop interface.

For a normal desktop PC, however, the standard LM Studio application with run server on login enabled is generally simpler.

Start the LM Studio Server From the Command Line

LM Studio includes the lms command-line utility with the desktop application. The CLI can manage models, the local API server, the inference runtime, and the headless daemon. 

Useful commands include:

lms server start

Start the local API server.

lms server stop

Stop the local server.

lms server status

Check whether the server is running.

lms ps

See which models are currently loaded.

lms ls

List the models available on your computer.

lms load <model>

Load a model manually.

LM Studio notes that you need to run the desktop application at least once before using the bundled lms CLI. 

These commands are useful if you want to automate your LM Studio environment using PowerShell, batch files, or Windows Task Scheduler.

Can LM Studio Automatically Load a Model at Startup?

You may not always need to load a specific model manually.

LM Studio supports Just-In-Time model loading, which allows API requests to load a downloaded model when it is needed rather than requiring you to preload every model during Windows startup. 

With JIT model loading enabled, a client can request an available model and LM Studio can load it into memory on demand.

This can be more convenient than permanently loading a large LLM into RAM or VRAM every time your PC starts.

If you do need a particular model loaded immediately, the CLI also supports:

lms load <model>

For example, a startup automation could start the LM Studio service and then load the model your applications normally use. The exact model identifier can be found with:

lms ls

Which Auto-Start Method Should You Use?

For most users, there are three practical setups.

If you use LM Studio normally but want the API available automatically:
Enable LM Studio's built-in option to run the LLM server on login.

If you want the LM Studio desktop interface to open every time you sign in:
Use Windows Startup Apps or the shell:startup folder.

If the computer is primarily a local AI server:
Consider using LM Studio's headless llmster daemon instead of relying on the desktop GUI.

The first option is usually the best balance for a normal Windows workstation.

LM Studio Keeps Starting With Windows: What Should You Check?

If LM Studio continues launching after you believe you disabled it, check both LM Studio and Windows.

First, open LM Studio Settings and make sure the run LLM server on login option is disabled.

Next, check:

Windows Settings > Apps > Startup

Then check:

Task Manager > Startup apps

Finally, inspect the current user's Startup folder by pressing Win + R and entering:

shell:startup

Windows can manage startup applications through more than one mechanism, so checking each location can help identify why an application continues to launch automatically. 

FAQ

How do I make LM Studio start automatically on Windows?

If your goal is to keep the local LLM API available, open LM Studio Settings and enable the option to run the LLM server when you log in. LM Studio can then operate in the background. 

How do I stop LM Studio from opening at startup?

Disable the server-on-login option inside LM Studio. Also check Settings > Apps > Startup or Task Manager > Startup apps in Windows and disable LM Studio there if necessary.

Can LM Studio run without the GUI?

Yes. LM Studio provides the llmster daemon for headless operation. It is designed to run independently of the desktop interface and is available for Windows. 

How do I start the LM Studio server automatically?

The easiest option on a desktop PC is LM Studio's server-on-login setting. You can also start the server manually or through automation with:

lms server start

Does LM Studio support OpenAI-compatible APIs?

Yes. LM Studio's local server includes OpenAI-compatible endpoints in addition to its native REST API and other supported interfaces. 

Do I need to load my model manually every time Windows starts?

Not necessarily. LM Studio supports Just-In-Time model loading, which can load downloaded models when API requests require them. You can also load a specific model with the lms load command.

Final Thoughts

Setting up LM Studio auto start on Windows depends on what you actually want to start.

If you simply want the LM Studio window to open after login, Windows Startup Apps can handle it.

If you want other applications to access your local LLM automatically, enable LM Studio's LLM server on login option instead. It keeps the server available in the background without requiring you to manually start it after every reboot.

And if you are turning a Windows machine into a dedicated local AI server, LM Studio's headless llmster daemon provides a cleaner server-oriented setup.

For most desktop users, the simplest configuration is:

Enable the LLM server on login, leave JIT model loading enabled, and let LM Studio run quietly in the background.