Repolode
Submit
lmstudio-ai

lmstudio-ai/lms

Command-line tool for LM Studio to manage local models, load and unload them, and control the local API server.

View on GitHubWebsite
Stars
5,363
Forks
495
Open issues
346
Last commit
6 days ago

What is it?

lms is the command-line interface for LM Studio, a desktop app for running large language models locally. It lets you check app status, list downloaded and loaded models, and load or unload them from the terminal, with options such as full GPU acceleration for loading.

The CLI also controls the local API server, so you can start and stop it from scripts or over SSH without opening the GUI. Output can be emitted as JSON for machine-readable use, and log streaming is available for debugging. It can scaffold a new project that uses the LM Studio SDK.

lms is built on lmstudio.js and is written in TypeScript on Node.js. It ships with LM Studio 0.2.22 and newer, and the repository notes that it is part of the lmstudio.js monorepo and cannot be built on its own.

Key features

  • 01Lists downloaded and loaded models, with JSON output for scripting.
  • 02Loads and unloads models, including GPU acceleration options.
  • 03Starts and stops the local API server from the terminal.
  • 04Streams LM Studio logs for debugging.
  • 05Creates new projects that use the LM Studio SDK.

Use cases

  • Managing local LLMs on a headless server or over SSH.
  • Automating model loading and server control in shell scripts or CI.
  • Bootstrapping a Node.js or TypeScript project that uses the LM Studio SDK.

Quick start

  1. 1

    Install LM Studio 0.2.22 or newer, which ships with lms

  2. 2

    If the command is not found, add lms to PATH using npx

    npx lmstudio install-cli
  3. 3

    Open a new terminal and check that lms runs

    lms
  4. 4

    Start the local API server

    lms server start

Things to know

  • Requires the LM Studio desktop app (0.2.22 or newer) to be installed.

Topics

#llm#cli#local-llm#typescript#nodejs#api-server#lmstudio

FAQ

What is lms?
Command-line tool for LM Studio to manage local models, load and unload them, and control the local API server. lms is the command-line interface for LM Studio, a desktop app for running large language models locally. It lets you check app status, list downloaded and loaded models, and load or unload them from the terminal, with options such as full GPU acceleration for loading.
Is lms free to use? What license does it have?
lms is open source under the MIT license. Check the license terms before using it in commercial products.
Is lms still maintained?
Actively maintained. The last commit to lms was on Oct 14, 2026.
What can I use lms for?
Managing local LLMs on a headless server or over SSH. Automating model loading and server control in shell scripts or CI. Bootstrapping a Node.js or TypeScript project that uses the LM Studio SDK.
How do I install lms?
Install LM Studio 0.2.22 or newer, which ships with lms If the command is not found, add lms to PATH using npx: npx lmstudio install-cli Open a new terminal and check that lms runs: lms Start the local API server: lms server start
Where can I get lms?
The source code and installation instructions are on GitHub: https://github.com/lmstudio-ai/lms

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects