Serve LLMs with LM Studio

Install llmster on the hardware platform

llmster is LM Studio's terminal-native, headless LM Studio daemon. Install it on servers, cloud instances, machines with no GUI, or your computer. This is useful for running LM Studio in headless mode on your hardware platform, then connecting from your laptop via the API.

On your hardware platform, install llmster by running:

curl -fsSL https://lmstudio.ai/install.sh | bash

Once installed, follow the instructions in your terminal output to add lms to your PATH. Interact with LM Studio using the lms CLI or the SDK / LM Studio V1 REST API (new with enhanced features) / OpenAI-compatible REST API.

Download required ancillary files

Run the following curl commands in your local (laptop) terminal to download files required for later steps. Choose Python, JavaScript, or Bash.

# JavaScript
curl -L -O https://raw.githubusercontent.com/lmstudio-ai/docs/main/_assets/nvidia-spark-playbook/js/run.js

# Python
curl -L -O https://raw.githubusercontent.com/lmstudio-ai/docs/main/_assets/nvidia-spark-playbook/py/run.py

# Bash
curl -L -O https://raw.githubusercontent.com/lmstudio-ai/docs/main/_assets/nvidia-spark-playbook/bash/run.sh

Open the script you downloaded and replace <SPARK_LOCAL_IP> with the actual IP address of your hardware platform. Step 3 shows how to find it. If you use LM Link (Step 4), use localhost instead. The script does not detect the IP address for you, and it fails with "LM Studio is not reachable" if the placeholder is left in.

Start the LM Studio API server

Use lms, LM Studio's CLI, to start the server from your terminal on the hardware platform. Enable local network access so devices on the same trusted local network can reach the API:

lms server start --bind 0.0.0.0 --port 1234

To test connectivity from your laptop, run:

curl http://<HARDWARE_IP>:1234/api/v1/models

where <HARDWARE_IP> is your hardware platform's reachable IP address. Find it on the hardware platform with:

hostname -I

(Optional) Connect with LM Link

LM Link lets you use models on your hardware platform from your laptop (or other devices) as if they were local, over an end-to-end encrypted connection. You don’t need to be on the same local network or bind the server to 0.0.0.0.

1. Create a Link — Go to lmstudio.ai/link and follow Create your Link to set up your private LM Link network.
2. Link both devices — On your hardware platform (llmster) and on your laptop, sign in and join the same Link. LM Link uses Tailscale mesh VPNs; devices communicate without opening ports to the internet.
3. Use remote models — On your laptop, open LM Studio (or use the local server). Remote models from your hardware platform appear in the model loader. Any tool that connects to localhost:1234 — including the LM Studio SDK, Codex, Claude Code, OpenCode, and the scripts in Step 7 — can use those models without changing the endpoint.

LM Link is in Preview and is free for up to 2 users, 5 devices each. For details and limits, see LM Link.

Download a model to your hardware platform

Download NVIDIA Nemotron 3.5 Super VL, the recommended model for this playbook, from the LM Studio catalog (nvidia/nemotron-3.5-super-vl):

lms get nvidia/nemotron-3.5-super-vl

This download can take a while due to its size. Verify the download:

lms ls

Load the model

Load the model on your hardware platform so it can respond to requests from your laptop:

lms load nvidia/nemotron-3.5-super-vl

Use the LM Studio SDK on the laptop

Install the LM Studio SDKs and use a simple script to send a prompt to your hardware platform and validate the response. Download the scripts from the Overview tab (or Step 2) and run the matching command from the directory that contains the script.

JavaScript

Prerequisites: npm and node installed

npm install @lmstudio/sdk
node run.js

Python

Prerequisites: uv installed

uv run --script run.py

Bash

Prerequisites: jq and curl installed

bash run.sh

Next steps

• Try downloading and serving different models from the LM Studio model catalog.
• Use LM Link to connect more devices and use models from your hardware platform from anywhere with end-to-end encryption.

Cleanup

Remove and uninstall LM Studio completely if needed. LM Studio stores models separately from the application. Uninstalling LM Studio does not remove downloaded models unless you explicitly delete them.

If you want to remove the entire LM Studio application on a GUI client, quit LM Studio from the tray first, then move the application to trash.

To uninstall llmster on the hardware platform, remove the folder ~/.lmstudio/llmster.

To remove downloaded models, delete the contents of ~/.lmstudio/models/.