lms CLI
LM Studio commands list downloaded and loaded models, control the local API server and stream application logs in a terminal.
- 5,404 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The CLI loads models into memory, uses GPU resources and changes local API server state.
Install
Manual install
npx lmstudio install-cliIf the command is missing from PATH, add it and open a new terminal.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The CLI loads models into memory, uses GPU resources and changes local API server state.
This is third-party code. Review the repository files before installing.
What it does
lms selects an LM Studio model, loads it for inference and starts the local API server. Separate commands report application status and models already loaded in memory. Unloading releases a model without deleting its downloaded file. Downloaded and loaded model lists support JSON for automation. A log stream helps investigate application behavior from a terminal. When several compatible applications are installed, the CLI selects an instance through its discovery order and uses that instance’s authentication and server settings.
Who it is for. Local AI application engineers managing LM Studio models from scripts or an agent terminal.
Good fit when
- An application author checks which model is ready for inference before a test.
- A test script starts the local API and stops it afterward.
- A script selects a model from a JSON list of downloaded files.
- During a generation failure, an engineer follows application logs in another terminal.
Not a fit when
- You need a standalone server without LM Studio or a compatible application.
- Your workflow treats a downloaded model as already loaded for inference.
- You cannot identify the intended instance among several running applications.
Example request
Show downloaded and loaded LM Studio models as JSON. Identify the selected application instance before loading anything new.Limitations
lms ships with LM Studio 0.2.22 and later and requires an installed compatible application. Among running apps, LM Studio and llmster take precedence over Bionic. If none are running, the CLI tries installations in the order llmster, LM Studio, Bionic. Loading a model needs device memory; `lms load` with `-y` skips confirmation and selects maximum GPU acceleration, so leave it off when first trying a model.
How to disable. Stop the API with `lms server stop`. Use `lms unload --all` to release loaded models; their downloaded files remain.
FAQ
How do I install lms?
Install [LM Studio](https://lmstudio.ai/), which includes the CLI. If the command is missing, run `npx lmstudio install-cli`, open a new terminal and run `lms`.
How do downloaded and loaded models differ?
`lms ls --json` lists downloaded files; `lms ps --json` lists models loaded for inference. Compare them before selecting a model.
Does checking status start the application?
`lms status`, `lms server status` and `lms server stop` do not start an application. Begin with `lms status` for a status-only check.
How do I start and stop the API?
Use `lms server start`, then `lms server stop` after the test. `lms server status` checks status without starting an application.
How do I load a model and observe its activity?
Use `lms load` to load a model. Run `lms log stream` in another terminal to follow logs; `lms ps` reports loaded models.
Discussion
No questions or reviews yet.
Related
CLI toolsLM Studio is a desktop app for running open models locally, with a chat, a local server in OpenAI and Anthropic formats, and MCP support.
A CLI and Python library for OpenAI, Claude, Gemini, local and other models: prompts, chats, tools, embeddings and a SQLite log
llmman runs Claude Code, Codex and OpenCode on a local or hosted model with one command and stores models as OCI images
A vLLM-style inference server on MLX that runs language models on Apple Silicon Macs and serves them over an API
A CLI connects a model to a Docker-container terminal so an agent can run commands, create files and save its step history
A local model server for Apple Silicon Macs in OpenAI, Anthropic and Ollama formats: Claude Code, Codex and OpenCode connect with one command
README in short
`lms create` starts a project using the LM Studio SDK. `lms --help` and `lms <subcommand> --help` document the installed release.