AutoML systems, which automate steps in model development and deployment process, have captivated the public imagination. Although robust end-to-end AutoML systems remain out of sight, Google recently open-sourced an AutoML framework for model architecture search at scale. Model Search enables researchers to apply many AutoML algorithms to automatically search for the best model architecture, ensemble models, or distilled models. Users can also create their own search space and compare models found during the search. While it could be extended to regression problems, it is designed to support classification problems.