Docker Model Runner has expanded its support to Windows developers by integrating vLLM on Docker Desktop for Windows with WSL2 and NVIDIA GPUs. Previously, vLLM was only available on Linux, limiting its accessibility. Now, Windows users can harness vLLM’s capabilities for high-throughput inference directly in their local environments. This is significant for developers building AI applications, as they can leverage their NVIDIA GPUs for faster performance. Docker Model Runner aims to simplify the process of running generative AI models. Users can execute models with a single command: `docker model run <model-name>`. The platform initially focused on NVIDIA GPUs but has expanded to include Vulkan support, making it compatible with a wider range of modern GPUs, including those from AMD and Intel. For users to get started, they need Docker Desktop 4.54 or later, WSL2 enabled, and an NVIDIA GPU with the appropriate driver updates. The setup involves enabling Docker Model Runner in Docker Desktop, installing the vLLM backend, and verifying that both inference engines are operational. Running models optimized for vLLM is straightforward, with users pulling models from Docker Hub that have a specific suffix. Troubleshooting memory issues is also covered, allowing users to adjust GPU memory utilization to accommodate multiple workloads. The update creates a more unified workflow for developers, allowing them to work in familiar environments while maintaining data privacy and reducing API costs during development. The article encourages community involvement through contributions and sharing, highlighting the collaborative nature of the Docker ecosystem.