LM Studio

LM Studio is a desktop application that runs large language models directly on a computer, rather than through a remote service. Users download models, load them into the application, and interact with them through chat. The application supports models from public model libraries and stores downloaded models on the local device. It is intended for users who want to work with language models without sending prompts to a cloud service. Downloaded models remain available on the device for offline use until removed.

Key Features

Local Models

Models run entirely on the local machine after download. No internet connection is required once a model is loaded, and prompts stay on the device.

Easy Discovery

The application includes a model browser for finding and downloading open models. A selected model can be loaded and used for chat after download completes.

Chat Interface

Talk to models in a chat interface. Test prompts and compare answers side by side.

Local API

Serve models through an OpenAI-like API on localhost. Build apps that use local AI instead of cloud services.

Voice Transcription

Speech can be transcribed offline in the application. Transcribed audio is processed on the device and is not sent to a remote service.

Frequently Asked Questions

What is LM Studio?

LM Studio runs AI models on your computer. You chat with them locally with full privacy and no internet.

Is LM Studio free?

Yes, for local use. The app is free to download and use for local models. Models are open weights you run yourself.

Do I need a powerful computer?

No, not necessarily. Smaller models run on modest machines. Pick a model size that fits your hardware.

Can I use it for my apps?

Yes. The local server mode gives you an API. Your apps can call local models like cloud ones.

Conclusion

LM Studio provides a way to run language models on a personal computer without a remote service. It supports model download, chat, prompt testing, and local application integration. Users choose a model size that matches their hardware, and smaller models can run on modest machines. That combination makes local language models accessible for study, development, and everyday tasks.