This article explains how to integrate Amazon SageMaker Model Cards with the model registry at the model version level for improved governance and documentation.

• Model cards now link to specific model versions in the registry, not just model names
• Enables standardized documentation across the entire model lifecycle
• Create model cards early in project lifecycle during business goal identification
• Update model cards during deployment phase with training and evaluation details
• Model card versions are immutable; changes create new versions automatically
• Clone previous model cards when iterating to new model versions
• Supports continuous ML documentation as part of MLOps best practices
• Python SDK enables linking model packages to model cards programmatically

This integration provides a single source of truth for model documentation, supporting governance, compliance, and accountability throughout the ML lifecycle.