Player not loading? Watch on YouTube
This tutorial walks through LM Studio setup on macOS, then downloads a local LLM and tests it in chat. The speaker describes LM Studio as a free desktop app that uses the computer's CPU or GPU and supports offline use. They present avoiding API charges and keeping prompts on the machine as reasons to run models locally; the walkthrough does not independently verify those privacy claims.
Installation starts at lmstudio.ai. The speaker opens the app download page, selects the Mac download and drags the downloaded app into the Applications folder. LM Studio opens without a model installed, so the next step is to use its model browser to find and download one.
The example uses a Mac mini with 16 GB of RAM. The transcript names Gemma 4 E4B, while the description names Gemma 3 E4B. That discrepancy leaves the exact model version uncertain. The speaker describes the selected download as roughly 6.86 GB and claims support for image input, reasoning and tool calling.
After downloading, the speaker selects the model from the chat dropdown and asks it to write a Python function to read a file. Ollama appears as a comparison for LM Studio's graphical interface. The description also promises a local OpenAI-compatible API server, but the transcript ends after the chat example and contains no server setup instructions.
