Skip to content


MLCube® is a project that reduces friction for machine learning by ensuring that models are easily portable and reproducible, e.g., between different stacks such as different clouds, between cloud and on-prem, etc.

Interested in getting started with MLCube? Follow the Getting Started instructions, or watch the video below.

Watch the video

Installing MLCube

Create python environment

conda create -n mlcube python=3.8
conda activate mlcube
virtualenv --python=3.8 .mlcube
source .mlcube/bin/activate

Install MLCube with docker and singularity runners

pip install mlcube mlcube-docker docker-singularity

Usage Examples

Check out the examples for detailed examples.


MLCube is licensed under the Apache License 2.0.

See LICENSE for more information.


Create a GitHub issue