Easiest way to run LLMs locally

You can even download, install and run models from the open-webui and keep a history, just like chatgpt.