I went ahead and created a dummy flow to test various combinations of providing parameters Code is below.

import pendulum
import prefect
from pendulum import DateTime
from prefect import flow
from pydantic import BaseModel, Field


class ExampleFlowParams(BaseModel):
    name: str = Field(description="The name of the birthday celebrator")
    dob: pendulum.Date = Field(description="Date of Birth")
    birthday_msg: str = "Happy birthday!",
    planned_birthday_party_date: DateTime | None = Field(description="When the party will happen, if any.", default=None)

    # @todo Can be removed once we update to Prefect 3.1.4 (that turns this on by default)
    class Config:
        arbitrary_types_allowed = True


@flow(name="Parameters Example Flow")
def parameters_example_flow(
    name: str,
    dob: pendulum.Date,
    params: ExampleFlowParams,
    planned_birthday_party_date: pendulum.DateTime | None = None,
    birthday_msg: str = "Happy birthday!",
):
    logger = prefect.get_run_logger()
    logger.info(f"{birthday_msg} {name} ({dob})!")
    if params.planned_birthday_party_date:
        logger.info(f"Looking forward to your party on {params.planned_birthday_party_date} or {planned_birthday_party_date}.")

You can see the result in the UI, a few things to highlight: • Seems like

DateTime

doesn't work either in a model or in the function arguments. Selecting the empty type label in the field results in an empty text box with

null

in the field being shown. • Date works as you've shown when it's defined as a function argument but not in a model. •

Happy new year! Any thoughts?