Docker Model Runner (DMR) pulls, runs and serves AI models locally. Models come from Docker Hub, other OCI registries and Hugging Face, and you can package GGUF and Safetensors files as OCI artifacts.
It exposes OpenAI- and Ollama-compatible APIs, runs on llama.cpp, vLLM and Diffusers inference engines, and can generate images with Stable Diffusion models.
Enable it in Docker Desktop, then use docker model commands, for example docker model configure --context-size 8192 ai/qwen2.5-coder.