You can enforce that with Pydantic validation. In Prefect 3.x (Pydantic v2), use

Field(min_length=3, max_length=4)

. If you’re defining this on a Pydantic model:

from typing import List
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: List[float] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2],
        min_length=3,
        max_length=4,
        description="",
        title="",
    )

If you want it directly as a flow parameter:

from typing import Annotated, List
from pydantic import Field
from prefect import flow

@flow
def my_flow(
    pesos: Annotated[List[float], Field(min_length=3, max_length=4)] = [0.4, 0.2, 0.2, 0.2]
):
    # your logic here
    return pesos

Both approaches only allow lists of length 3 or 4; anything else will raise a validation error. Note: If you happen to be on Prefect 2.x (Pydantic v1), use `min_items`/`max_items` or

conlist

, e.g.:

from pydantic import conlist, BaseModel

class Params(BaseModel):
    pesos: conlist(float, min_items=3, max_items=4) = [0.4, 0.2, 0.2, 0.2]

Got it — quick clarifier: are you trying to enforce - the list’s length (min number of items), or - the min/max value for each float in the list? Also, which Pydantic version are you on? Prefect 3.x uses Pydantic 2.x, and the syntax differs from 1.x. If you mean list length: - Pydantic 2.x (Prefect 3.x): use

min_length

/

max_length

from typing import List
from typing_extensions import Annotated
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: Annotated[List[float], Field(min_length=3, max_length=4)] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2]
    )

# This should raise a ValidationError (too few items)
# Params(pesos=[0.1, 0.2])

- Pydantic 1.x: use

min_items

/

max_items

(or

conlist

)

from typing import List
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: List[float] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2],
        min_items=3,
        max_items=4,
    )

or

from pydantic import BaseModel, conlist

class Params(BaseModel):
    pesos: conlist(float, min_items=3, max_items=4) = [0.4, 0.2, 0.2, 0.2]

If you mean per-item bounds (e.g., each weight between 0 and 1): - Pydantic 2.x:

from typing import List
from typing_extensions import Annotated
from pydantic import BaseModel, Field

Number = Annotated[float, Field(ge=0, le=1)]

class Params(BaseModel):
    pesos: Annotated[List[Number], Field(min_length=3, max_length=4)] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2]
    )

- Pydantic 1.x:

from pydantic import BaseModel, conlist, confloat

Weight = confloat(ge=0, le=1)

class Params(BaseModel):
    pesos: conlist(Weight, min_items=3, max_items=4) = [0.4, 0.2, 0.2, 0.2]

If you’re still seeing “max works but min doesn’t,” please share: - your Pydantic version - an example input that should fail the min check but doesn’t I can reproduce and pinpoint the issue from there.

@Marvin i am truin g to fix the min number of items to be 3 and the maximum to be 4

Thanks! To constrain the list length to min=3 and max=4: First, which Pydantic version are you on? Prefect 3.x uses Pydantic 2.x. Use the matching snippet below. Pydantic 2.x (Prefect 3.x) — use min_length/max_length

from typing import List
from typing_extensions import Annotated
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: Annotated[List[float], Field(min_length=3, max_length=4)] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2]
    )

# Fails (too few items)
# Params(pesos=[0.1, 0.2])

# Fails (too many items)
# Params(pesos=[0.1, 0.2, 0.3, 0.4, 0.5])

# OK (3 or 4 items)
# Params(pesos=[0.1, 0.2, 0.3])

Pydantic 1.x — use min_items/max_items or conlist

from typing import List
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: List[float] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2],
        min_items=3,
        max_items=4,
    )

or

from pydantic import BaseModel, conlist

class Params(BaseModel):
    pesos: conlist(float, min_items=3, max_items=4) = [0.4, 0.2, 0.2, 0.2]

Common gotcha: if you provide a default of length 4 (via

default_factory

), omitting

pesos

will always pass. To actually see the “min items” error, pass a shorter list:

Params(pesos=[0.1, 0.2])  # should raise a ValidationError

If you still see “max works but min doesn’t,” please share your Pydantic version and the exact input that isn’t failing so I can reproduce.

Got it. Two quick points to make sure this works end-to-end: 1) Pydantic constraints - If you’re on Pydantic 2.x (Prefect 3.x), use min_length/max_length on the field. - If you’re on Pydantic 1.x (Prefect 2.x), use min_items/max_items (or conlist). Pydantic 2.x (Prefect 3.x)

from typing import List
from typing_extensions import Annotated
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: Annotated[List[float], Field(min_length=3, max_length=4)] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2]
    )

Pydantic 1.x

from typing import List
from pydantic import BaseModel, Field

class Params(BaseModel):
    pesos: List[float] = Field(
        default_factory=lambda: [0.4, 0.2, 0.2, 0.2],
        min_items=3,
        max_items=4,
    )
# or
from pydantic import BaseModel, conlist

class Params(BaseModel):
    pesos: conlist(float, min_items=3, max_items=4) = [0.4, 0.2, 0.2, 0.2]

2) Prefect UI vs server-side enforcement - The Prefect 3 UI may allow you to enter 2 items in the form, but the server will reject the run at creation if

enforce_parameter_schema

is enabled on the deployment (true by default) and your flow parameter is typed with your

Params

model. - Make sure your flow signature actually uses the model so the schema is generated:

from prefect import flow

@flow
def my_flow(params: Params):
    ...

If you still see a run start with only 2 items (no validation error), it usually means one of: - The flow parameter is not typed as

Params

(so the deployment has no schema for it) - You’re on Pydantic 1.x but using

min_length

instead of

min_items

- Schema enforcement is disabled for the deployment/run If you can share: - Prefect version (

prefect version

) and Pydantic version (

import pydantic; pydantic.__version__

) - Your flow signature - How you’re starting the run (UI/CLI/API) …I can pinpoint exactly why the 2-item case is slipping through.