What is the difference between basemodel and pydantic_model_creator?

i'm a beginner in fastapi; i have a question about returning the contents of api functions.

i have knew that BaseModel could regulate the input parameters of Api functions; i used the tortoise orm in fastapi and i used "pydantic_model_creator" function to serializer returning the contents of api functions.

but,recnetly i saw another use of basemodel; it was used to define the returning contents of api functions.

example: a function about use model;

base.py：

class UserInfo(BaseModel):
    id : int 
    user_name: str
    age: Optional[int]
    user_type: bool
    nickname: Optional[str]
    user_phone: Optional[str]
    user_email: Optional[str]
    full_name: Optional[str]
    user_status: bool
    header_img: Optional[str]
    sex: int

serializer.py:

single_user =  pydantic_model_creator(User, name="single_user", exclude=("password", "remarks", "client_host", "create_time", "update_time",))

fuunction.py:

async def user_info(req: Request):
    """
    get current user info
    user_id: int
    """
    user_data = await User.get_or_none(pk=req.state.user_id)
    if not user_data:
        return fail(msg=f"ID{req.state.user_id} is fail!")
    print(user_data.__dict__)
    # return success(msg="user info", data=UserInfo(**user_data.__dict__))
    return await single_user.from_tortoise_orm(user_data) 

So, I want a suggestion. Which of the two uses is better?

There is no winner in your question. The two serve two different purposes.

What is a BaseModel?

A Pydantic BaseModel is a class that defines how your data looks like and the validation requirements it needs to pass in order to be valid.

For further information see https://docs.pydantic.dev/latest/

What's the use of pydantic_model_creator?

pydantic_model_creator is a function from the library tortoise-orm. It is part of this library and thus thought for being used with it. The goal is to transform the declared ORM model into a pydantic model that works with other web frameworks (e.g. pydantic models can be used also with django).

The big picture: how to use them?

You can use tortoise-orm as ORM and then declare separately all the single pydantic models. This will duplicate the effort required and increase the coupling between the different components of your project. It is thus advisable to use pydantic_model_creator to create the input models that your API will consume and feed them to the ORM (the converse also applies since you'll retrieve data and return it to the user).

The only case in which it would make sense to use BaseModel instead of pydantic_model_creator is if you need to return a particular set of fields or data structured in a particular way. Though, this should be an exception/rare case and not the norm.

On the other hand, if you're not using tortoise-orm, you'll need to write the models (BaseModel) yourself or use the tools provided by the ORM.