LM Studio

Desktop app for downloading and chatting with local models, with a local OpenAI-compatible server and the lms command line. Suits a workstation more than a headless server; for servers prefer Ollama or llama.cpp.

Project: LM Studio Documentation: LM Studio docs · lms CLI

Verification: the whole sheet is (unverified): LM Studio was not installed here and the commands are from the vendor docs as remembered. Run lms --help to confirm.

CLI (`lms`) #

COMMAND	DESCRIPTION
`lms status`	Show server and model state
`lms ls`	List downloaded models
`lms get MODEL`	Download a model
`lms load MODEL`	Load a model into memory
`lms unload --all`	Unload all models
`lms ps`	Show loaded models
`lms server start`	Start the local API server
`lms server stop`	Stop it

Local Server #

The server listens on port 1234 by default and speaks the OpenAI API. See OpenAI API.

bash

curl http://localhost:1234/v1/models

Notes #

Models are GGUF (and MLX on Apple silicon). See Hugging Face CLI and Model Formats.
Bind the server to the network only if you need to, and firewall it. It has no authentication by default (unverified).

See Also #

Updated 9 Nov 2026 Edit this page on GitHub