Local runtime - Docker - verified
Docker Model Runner: install, API, platforms and features
Docker Model Runner is a local LLM runtime from Docker. As its docs describe it: Docker Model Runner is for developers who want to pull, run and serve AI models with Docker tooling from Docker Hub, any OCI-compliant registry or Hugging Face. The latest release is v1.2.8, published August 12, 2026. License: Apache-2.0 (open source).
- Latest release
- v1.2.8
- Released
- 2026-08-12
- License
- Open source
- OpenAI API
- Yes
Install
Copied verbatim from the page linked under each command. Labels are worded as the source words them.
linux (Ubuntu/Debian, Docker Engine)
sudo apt-get install docker-model-pluginlinux (RPM-based, Docker Engine)
sudo dnf install docker-model-pluginmacos (arm64, standalone dmr)
brew install docker/tap/dmrSource: github.com/docker/model-runner
windows (amd64, standalone dmr)
winget install Docker.dmrSource: github.com/docker/model-runner
Run a model
The first-run example from the official docs. OpenAI-compatible endpoint: http://localhost:12434/engines/v1
docker model run ai/smollm2Features
What the Docker Model Runner docs say it supports. Not stated means the docs do not say either way.
| Feature | Value | Source |
|---|---|---|
| Interface | CLI, Desktop app, Server, Docker | See sources below |
| Platforms | macOS, Windows, Linux | docs.docker.com/ai/model-runner |
| Backends | CPU, CUDA, ROCm, Metal, Vulkan, Other | docs.docker.com/ai/model-runner/inference-engines |
| Model formats | GGUF, Safetensors, Other | docs.docker.com/ai/model-runner/inference-engines |
| OpenAI-compatible API | Yeshttp://localhost:12434/engines/v1 | docs.docker.com/ai/model-runner/api-reference |
| Anthropic-compatible API | Yes | docs.docker.com/ai/model-runner/api-reference |
| Tool calling | Yes | docs.docker.com/ai/model-runner/api-reference |
| Structured outputs | Yes | docs.docker.com/ai/model-runner/api-reference |
| Vision (image input) | Yes | docs.docker.com/ai/model-runner/api-reference |
| Multi-GPU | Yes | docs.docker.com/ai/model-runner/inference-engines |
| MCP client | Not stated | Docs do not say |
| License | Apache-2.0Open source | github.com/docker/model-runner/blob/main/LICENSE |
| Latest release | v1.2.8August 12, 2026 | github.com/docker/model-runner/releases/tag/v1.2.8 |
Models to run with Docker Model Runner
Browse the open-weight models you can run locally for download sizes, licenses and memory fit; check which formats Docker Model Runner loads in the table above.
Compare Docker Model Runner
FAQ
Does Docker Model Runner have an OpenAI-compatible API?
Yes. Endpoint, per its docs: http://localhost:12434/engines/v1. Source: https://docs.docker.com/ai/model-runner/api-reference/
Does Docker Model Runner have an Anthropic-compatible API?
Yes, per https://docs.docker.com/ai/model-runner/api-reference/
Which platforms and GPUs does Docker Model Runner support?
Its docs list macOS, Windows and Linux, with these backends: CPU, CUDA, ROCm, Metal, Vulkan and Other.
How do I install Docker Model Runner?
On linux (Ubuntu/Debian, Docker Engine): "sudo apt-get install docker-model-plugin", as shown at https://docs.docker.com/ai/model-runner/get-started/. This page lists 4 install commands: linux (Ubuntu/Debian, Docker Engine), linux (RPM-based, Docker Engine), macos (arm64, standalone dmr) and windows (amd64, standalone dmr).
Is Docker Model Runner open source?
Yes. It is released under Apache-2.0; the license is at https://github.com/docker/model-runner/blob/main/LICENSE.
What is the latest version of Docker Model Runner?
v1.2.8, published August 12, 2026 (https://github.com/docker/model-runner/releases/tag/v1.2.8), as of October 18, 2026.
Which model formats does Docker Model Runner run?
GGUF, Safetensors and Other, per https://docs.docker.com/ai/model-runner/inference-engines/
Sources
Last verified from the pages below.
- Docker Model Runner docs - docs.docker.com
- Get started with DMR - docs.docker.com
- DMR REST API reference - docs.docker.com
- Inference engines - docs.docker.com
- docker/model-runner README - github.com
- Release v1.2.8 - github.com
- Releases Atom feed - github.com
Newsletter
Local AI, every week
One email per week. New model releases, runtime updates and what they mean for what you build.
Keep exploring
Open-weight models you can run
Download sizes, licenses and the run command for each model.
All local LLM runtimes
Every runtime in one feature matrix, with sources.
Picking a runtime for coding agents
Ollama, LM Studio, vLLM and llama.cpp for agent workloads.
Setup guides
Point your coding tools at the model you picked.