One of the most exciting features of AI and its components — machine learning, deep learning, edge computing — is that it allows multidisciplinary research teams to work together to accelerate the pace of discovery. The proliferation of AI methods across domains has resulted in an explosion in journal publications and creation of new models and codes. However, finding and using the latest models in new applications is often challenging. To overcome this challenge, Argonne has developed a state-of-the-art repository to collect, publish,and categorize models, making them easily accessible to others who want to use them. The repository is operated as a service for researchers to simplify model sharing and use. The system uses unique identifiers and versioning and assigns metadata to make models easily discoverable.