Docker Model Runner
Docker Model Runner is a local AI model execution component provided by the Docker platform, enabling developers to pull, manage, and run large language models and other AI models using standard Docker commands. It containerizes the model runtime environment and provides an OpenAI API-compatible interface, allowing developers to integrate and test AI model capabilities in local environments without relying on external cloud services.
Timeline (last 90 days)
Docker Agent adopts a model-agnostic strategy, supporting cloud models such as OpenAI, Anthropic, Gemini, xAI, and Mistral, and can also run local models via Docker Model Runner.
Docker Agent is compatible with OpenAI, Anthropic, Gemini, and Docker Model Runner, and is not locked to a single vendor.
Tools such as LM Studio, Ollama, and Docker Model Runner are essentially wrappers around inference engines.
Docker Model Runner is currently available as an experimental feature of Docker Desktop, performs well on Linux with NVIDIA hardware, and exposes a REST API port at approximately 12434.
All Facts (7)
容器内运行时通过model-runner.docker.internal这个内部DNS与Docker Model Runner模型服务通信
90%Unverified本地开发时Docker Model Runner的模型API地址为localhost:12434
85%UnverifiedDocker Model Runner使用GGUF格式的量化模型文件，这些文件被打包为符合OCI规范的artifact
85%UnverifiedDocker Agent adopts a model-agnostic strategy, supporting cloud models such as OpenAI, Anthropic, Gemini, xAI, and Mistral, and can also run local models via Docker Model Runner.
50%UnverifiedDocker Agent is compatible with OpenAI, Anthropic, Gemini, and Docker Model Runner, and is not locked to a single vendor.
50%UnverifiedTools such as LM Studio, Ollama, and Docker Model Runner are essentially wrappers around inference engines.
50%UnverifiedDocker Model Runner is currently available as an experimental feature of Docker Desktop, performs well on Linux with NVIDIA hardware, and exposes a REST API port at approximately 12434.
50%