A local LLM is a language model that runs on hardware you control. After downloading the model, prompts can be processed on your machine rather than sent to a cloud API.
In practice
Local LLMs are useful for privacy-sensitive experiments, offline work, cost control, learning, and custom internal tools. They are often smaller or quantized to fit consumer hardware.
What to watch
Local does not always mean safer or better. You still need secure storage, good model choices, and enough hardware for acceptable speed.