Here’s my setup simplified:

from abc import ABC
from typing import TypeVar
from pydantic import BaseModel

_M = TypeVar("_M", bound=BaseModel)

class A(BaseModel):
    a = 1
    b = 2
    c = 3

class Base(ABC):
    cfg_model: type[_M]

    def __init__(self):
        self.cfg = self.cfg_model()

class ChildA(Base):
    cfg_model = A
    
    def run(self):
        self.cfg_model  # (variable) model_class: type[_M]
        self.cfg        # (variable) cfg: _M

        # pylance fails to detect properties in `self.cfg`
        self.cfg.a
        self.cfg.b
        self.cfg.c

The idea is that there are multiple classes extending the base abstract class, each having it’s own config model.

But i can’t seem to figure out how to make pylance (i’m using Visual Studio Code) correctly typecheck self.cfg in ChildA‘s run method as A.

UPD: Turning Base into generic class helped:

from abc import ABC
from typing import TypeVar
from pydantic import BaseModel

_M = TypeVar("_M", bound=BaseModel)

class A(BaseModel):
    a = 1
    b = 2
    c = 3

class Base(ABC, Generic[_M]):
    cfg_model: type[_M]

    def __init__(self):
        self.cfg = self.cfg_model()

class ChildA(Base[A]):
    cfg_model = A
    
    def run(self):
        self.cfg_model  # (variable) model_class: type[A]
        self.cfg        # (variable) cfg: A

        # all properties are correctly typechecked
        self.cfg.a
        self.cfg.b
        self.cfg.c

However, it irks me that i need to provide the Model name twice (in class argument and in class variable). Can this be avoided somehow?