A practical guide to running large language models locally on your own hardware — covering Ollama, LM Studio, hardware requirements, and when local beats cloud.