Keep updates type-safe
Mutable models whose fields may change after initialization.
Pydantic · Pipe
validate_assignment
validate_assignment enables validation when fields are changed after a Pydantic model is created. It defaults to False, so post-construction assignments are otherwise not revalidated.
How it works
Set validate_assignment=True in a model’s configuration to validate values assigned after initialization. Pydantic then applies validation when model attributes change, rather than validating only during model creation.
Syntax
class User(BaseModel, validate_assignment=True):
...
# or
model_config = ConfigDict(validate_assignment=True)Key points
validate_assignmentcontrols whether Pydantic checks values assigned after the model has been created.- The option is disabled unless you explicitly configure it.
- Initial model construction is validated independently of assignment-validation configuration.
- Without this option, mutating a model does not trigger another validation pass.
- The configuration can be supplied through the model class declaration or through
model_config. - When enabled, later attribute changes receive validation in addition to construction-time checks.
Common mistakes
- Assignment validation is disabled by default, so changing an attribute otherwise bypasses model validation.
Example: Validate assignments after model creation
from pydantic import BaseModel, ValidationError
class User(BaseModel, validate_assignment=True):
name: str
age: int
user = User(name="Ada", age=36)
user.age = "37"
assert user.age == 37
try:
user.age = "not an integer"
except ValidationError:
print("assignment rejected")The first assignment is converted to an integer; the second raises a validation error.
When to use it
Preserve validation guarantees
Applications that modify validated state over time.
Reject invalid mutations
Models where accidental invalid assignments should fail immediately.
Interview questions
QWhat does validate_assignment change?
By default, Pydantic validates data when the model is created, but not when an attribute is changed. Enabling validate_assignment makes later assignments undergo validation too.
Follow-up: How can it be enabled? Use validate_assignment=True as a model class keyword argument or set it in model_config.
QHow do you enable assignment validation?
It can be enabled directly in the class definition with class User(BaseModel, validate_assignment=True). This causes assignments such as user.age = ... to be validated after construction.
Follow-up: What happens without the setting? An invalid value can be assigned without revalidation.
QWhat is the default behavior?
The setting is a configuration option whose default is False. With the default, validation occurs only during model creation; with it enabled, validation also occurs when model data changes.
Follow-up: Why might this matter? It preserves the model’s validation guarantees across its full mutable lifetime.
Related
BaseModelConfigDictmodel_configValidationError