Exclude one or more fields in Pydantic model from appearing in FastAPI responses

The Problem

Some of the Pydantic models in my FastAPI project include secret internal fields that I don’t want to expose in the API response or generated documentation. For example, I would like to exclude the fields address and created_at in the User model below:

from pydantic import BaseModel
from datetime import datetime

class User(BaseModel):
    username: str
    email: str
    address: str
    created_at: datetime

How do I keep these fields from appearing in responses while keeping them in the model?

The Solution

To exclude multiple fields from a Pydantic model, we can expand the type definition using Annotated from Python’s built-in typing module. This special typing form was proposed in PEP 593 and is used to add specific metadata to type declarations. This metadata is being used by libraries such as Pydantic to enable special behavior. In this case, we’ll use it to exclude the address and created_at fields from our API’s responses:

from pydantic import BaseModel, Field # also import Field
from datetime import datetime
from typing import Annotated # import Annotated

class User(BaseModel):
    username: str
    email: str
    address: Annotated[str, Field(exclude=True)]
    created_at: Annotated[datetime, Field(exclude=True)]

Note that all of Pydantic’s standard validation features will still work for excluded fields.