Docker Model Runner (DMR) is a new feature in Docker Desktop that enables running open-weighted LLMs locally, similar to Ollama. This tutorial shows how to call DMR from Python using the OpenAI API Python SDK. No prior Docker knowledge is required (operate as a server)