LM Studio is a desktop application for running large language models locally, used to download and run models such as Qwen, Mistral, Gemma and gpt-oss. The documented flow is to install the app, download a model from the Discover tab or search for one, then load it in the Chat tab, where a model loader selects a downloaded or sideloaded model and optionally sets load configuration parameters before the conversation begins; loading a model allocates memory for its weights and other parameters. It also ships Bionic, an agent for work and code that creates and edits documents, slides, PDFs and software, handles coding tasks, automations and computer control, and transcribes speech in real time with voice and audio data processed locally. The app is powered by the LM Studio runtime with MLX and llama.cpp under the hood, can run the latest frontier open models, and its cloud services operate under zero data retention.
LM Studio
Latest version:Unknown
Release date:Unknown
A desktop app for running large language models on your own machine, with a Discover tab for downloading models and a Chat tab for loading and conversing with them. It includes an agent for documents, slides, PDFs and code work, plus real-time voice transcription processed locally.