Player not loading? Watch on YouTube
This tutorial walks through installing LM Studio on Windows and using its desktop interface to run models locally. The setup starts at lmstudio.ai: download the Windows installer, open the executable from the Downloads folder, and follow the prompts. The speaker says the installer handles setup dependencies automatically.
The model search uses the welcome screen's search bar or the Discover tab. Llama 3 and Qwen are example searches, with community-quantized GGUF variants appearing in the results. For a computer with 16 GB of RAM, the tutorial recommends a smaller 3B or 7B model. It gives a starting point for model selection without testing those sizes on specific hardware.
After the download, the walkthrough opens Chat and selects the model from the loading menu at the top. LM Studio loads it into system memory. Users with a dedicated graphics card can adjust the right-hand hardware settings to offload model layers to GPU VRAM; the speaker presents this as a way to speed up generation but provides no benchmark.
The final step is entering a prompt in the chat box. The speaker describes responses as offline and private after the model has downloaded and loaded, without cloud API calls. The guide focuses on desktop setup and basic chat.
