here's my attempt at unifying SQLAlchemy and Pydantic. I use the new dataclasses support in both libraries to do this. https://gist.github.com/sandys/671b8b86b

here's my attempt at unifying SQLAlchemy and Pydantic. I use the new dataclasses support in both libraries to do this. https://gist.github.com/sandys/671b8b86ba913e6436d4cb22d04b135f https://gist.github.com/sandys/671b8b86ba913e6436d4cb22d04b1... This example trys to play with some quirky table structures - Mixins, Surrogate primary keys, index, etc Both async and sync modes. I like what @tiangolo has done with SQLModel, but i suspect that a lot of people who are already using sqlalchemy in production will prefer to unify via dataclasses versus switching the DX to a new library. And this is possible today.