Mypy handling initializing Pydantic models with dict values incorrectly

[dpitch40]

Bug Report

Mypy is flagging code pasted directly from the Pydantic docs (involving initializing models with dictionary values for nested classes) as incorrectly typed.

To Reproduce

Save the following code (directly from the current Pydantic docs to a Python file.

from typing import Literal, Union

from pydantic import BaseModel, Field, ValidationError


class Cat(BaseModel):
    pet_type: Literal['cat']
    meows: int


class Dog(BaseModel):
    pet_type: Literal['dog']
    barks: float


class Lizard(BaseModel):
    pet_type: Literal['reptile', 'lizard']
    scales: bool


class Model(BaseModel):
    pet: Union[Cat, Dog, Lizard] = Field(..., discriminator='pet_type')
    n: int


print(Model(pet={'pet_type': 'dog', 'barks': 3.14}, n=1))

Then run mypy on it.

Expected Behavior

Mypy should validate the above code without errors, as it is example code directly from Pydantic's docs.

Actual Behavior

test.py:26: error: Argument "pet" to "Model" has incompatible type "dict[str, object]"; expected "Cat | Dog | Lizard"  [arg-type]
Found 1 error in 1 file (checked 1 source file

Your Environment

• Mypy version used: 1.10.0
• Mypy command-line flags: None
• Mypy configuration options from mypy.ini (and other config files): mypy.ini
• Python version used: 3.10.11
• Pydantic version used: 2.7.1

[dpitch40]

I can also reproduce it with this code from the Pydantic docs.

from typing import List, Optional

from pydantic import BaseModel


class Foo(BaseModel):
    count: int
    size: Optional[float] = None


class Bar(BaseModel):
    apple: str = 'x'
    banana: str = 'y'


class Spam(BaseModel):
    foo: Foo
    bars: List[Bar]


m = Spam(foo={'count': 4}, bars=[{'apple': 'x1'}, {'apple': 'x2'}])

The Pydantic docs appear to treat code initializing nested classes using dictionaries containing the fields and values as correct (and the code does work for me if I run it), but Mypy is flagging such code as incorrect.

[hauntsaninja]

mypy isn't going to support this. I think pydantic has a mypy plugin that might help with this, you could ask on their repo.