False positive - "Parameter unfilled" since Pycharm 2024.3

[UltimateLobster]

Describe the bug
Ever since upgrading to Pycharm 2024.3, fields declared with an Annotated style do not have their default_factory argument recognized. Calling the model's constructor will report an error if I don't provide a value to the field, although it has a default factory.

To Reproduce
Steps to reproduce the behavior:

1. Install Pycharm 2024.3+
2. Create a model with a field declared with the Annotated style.
3. Add a default factory to the field
4. Create an instance of the model without any arguments

Expected behavior
No error or warnings should be shown. Although we didn't pass a value to our field, it has a default factory so it's fine.

Screenshots

Environments (please complete the following information):

• IDE: Pycharm Professional - 2024.3
• OS: Windows 11
• Pydantic Version: 2.9.2
• Plugin Version: 0.4.6

Additional context
Pycharm 2024.3 added support for dataclass_transform classes (like pydantic.BaseModel) which seems to have caused a false-positive for models which declare fields with a default factory using Annotated style.

I couldn't find anything in the spec mentioning the use of Annotated, so it seems appropriate from Pycharm's side of things, to not recognize default_factory in this way.

However, something about their way of implementing may have overridden this plugin's previous behavior.

[primal100]

Probably related, annotated types give incorrect type errors in 2024.3:

from pydantic import BaseModel, EmailStr


class CustomModel(BaseModel):
    email: EmailStr
    variable: int = 1

c = CustomModel(email="a@b.com", variable=1)

[jzr-supove]

This is still not resolved, any workarounds except adding # noqa next to it?

[mrpiggi]

This holds for trivial classes as well as soon as you import them from another module. So defining

class CustomModel(BaseModel):
    variable: int = 1

in model.py and using it in logic.py with

from model import CustomModel

custom_model = CustomModel()

complains about missing required arguments.

(PyCharm 2025.2.0.1Build #PY-252.23892.515 and Pydantic Plugin 0.4.19)

See #1086

[krmozejko]

I experience the same behavior as described by mrpiggi, PyCharm 2025.2.2 Build #PY-252.26199.168 with Pydantic Plugin 0.4.19.