dbt Models Documentation (2026)

Official dbt documentation for models — last updated May 18, 2026.

Key facts surfaced (2026-05-19):

• Models are SQL (or Python in dbt 1.3+) SELECT statements that transform warehouse data without data leaving the system.
• Run via dbt run; materialize as tables or views in the target data warehouse.
• Models reference each other using ref() function to build dependency DAGs.
• dbt handles the "T" (Transform) in ELT; data is loaded into the warehouse before transformation.
• Curriculum reference: composable CDP equivalent of AEP Data Prep mapping functions for data transformation (Modules 2-3).