**Ollama** is an open-source large language model service that provides an API interface and chat interface similar to OpenAI. It allows seamless deployment of the latest GPT models and usage through APIs. It supports hot-loading model files, enabling model switching without restarting.