dbt-pydantic-generator

To generate Pydantic models from dbt schemas for Airflow data validation, run the generator against your dbt model path to produce ready-to-use BaseModel classes. It automatically infers column names, data types, and nullability to enforce type-safe data contracts in DAGs.

Yes, you can handle Snowflake VARIANT JSON fields because the generator infers JSON and VARIANT column types from dbt models. It provides built-in helper methods like to_snowflake_dict and from_snowflake_row for seamless Airflow integration.

The best way to enforce data contracts across dbt model layers in Airflow is generating strongly typed Pydantic models. The generator applies type inference and nullability handling across raw, base, clean, intermediate, fact, and dimension layers to enable runtime validation.

No additional dependencies are required to use dbt-pydantic-generator for type-safe DAGs. It operates standalone with no listed dependencies, directly converting dbt SQL schemas into Pydantic BaseModel classes with generated validators for IDE support.

Yes, the generated Pydantic model handles nullability for dbt column types. It implements column type inference and nullability handling from dbt schema outputs, ensuring runtime data validation correctly processes optional fields within Airflow DAGs.