Central repository for storing, versioning, and managing trained machine learning models with their metadata. Like a library catalog system that tracks all versions of AI models.
A model registry stores each version of a recommendation model, tracking which version is in production, staging, or archived.
All provide a centralized place to register trained models, track versions and metadata, and manage lifecycle states (e.g., staging/production). AWS and GCP emphasize registry + deployment workflows; Azure uses registries/workspaces to organize and govern models across teams; OCI provides a model catalog for storing and managing model artifacts and metadata within OCI Data Science.