Absolutely — here are two complete Prefect 3.x examples you can drop into a Python script to define and create a deployment. Example A: Deploy a locally defined flow (references this file as the entrypoint)

from prefect import flow, get_run_logger

@flow
def etl(name: str = "world"):
    logger = get_run_logger()
    <http://logger.info|logger.info>(f"Hello, {name}!")

if __name__ == "__main__":
    # Creates/updates a deployment for this flow
    deployment_id = etl.deploy(
        name="etl-dev",
        work_pool_name="my-work-pool",        # replace with your work pool
        description="Example deployment defined in Python",
        parameters={"name": "Prefect"},
        cron="0 * * * *",                     # run hourly; or use interval=60 or remove for ad-hoc
        tags=["example"],
        # If your work pool uses containers (Docker, K8s, ECS), set an image:
        # image="<http://ghcr.io/your-org/your-image:latest|ghcr.io/your-org/your-image:latest>",
        # job_variables={}  # pool-specific overrides; set as needed for your worker type
    )
    print(f"Created deployment: {deployment_id}")

Example B: Deploy a flow from a GitHub repo (code-based deployment from source)

from prefect import flow
from prefect.flows import Flow

@flow
def etl(name: str = "world"):
    # Flow body can be minimal; actual code will be read from the repo at runtime
    pass

if __name__ == "__main__":
    Flow.from_source(
        source="<https://github.com/your-org/your-repo>",
        entrypoint="flows/etl.py:etl",        # path within repo + ':' + flow function name
    ).deploy(
        name="etl-prod",
        work_pool_name="k8s-pool",            # replace with your work pool
        image="<http://ghcr.io/your-org/your-image:latest|ghcr.io/your-org/your-image:latest>",
        parameters={"name": "Production"},
        cron="0 0 * * *",                     # run daily at midnight
        tags=["prod"],
        # job_variables={}  # optional, depends on your work pool (e.g., env, resources)
    )

Notes - This assumes Prefect 3.x. In 3.x, use

Flow.from_source(...).deploy(...)

for repo-based deployments and

my_flow.deploy(...)

for local code. The deprecated 2.x APIs like

Deployment.build_from_flow()

and

prefect deployment build

are removed. - Make sure you have a worker running for the target work pool so scheduled and ad-hoc runs get picked up. -

job_variables

are work-pool specific — set them to match your pool’s job template (e.g., env, resources, volumes). Docs and references - Create deployments via Python: Deploy via Python - Deployment how-to overview: Create deployments - REST/API reference for deployments: Create deployment API Want me to tailor this to your setup? Share: - Your work pool type/name (e.g., local, Docker, Kubernetes, ECS) - Whether your flow code lives locally or in a Git repo - Any schedule and default parameters you need

Got it — if you mean the Prefect “Deployment” class, quick check: are you on Prefect 2.x? In Prefect 3.x there isn’t a public

Deployment

class; the 3.x equivalent is

RunnerDeployment

plus the

deploy(...)

helper. Below are complete 3.x scripts using that pattern. If you really are on 2.x and want the 2.x

Deployment

class example, say so and I’ll send it. Prefect 3.x: code-based deployment with RunnerDeployment (local entrypoint)

from prefect import flow
from prefect.deployments.runner import RunnerDeployment
from prefect.deployments import deploy
from prefect.deployments.schedules import create_deployment_schedule_create
from prefect.schedules import Cron

@flow
def etl(name: str = "world"):
    print(f"Hello, {name}!")

if __name__ == "__main__":
    d = RunnerDeployment(
        name="etl-dev",
        flow_name="etl",  # Must match the @flow function name
        entrypoint="path/to/your_file.py:etl",  # Update to your file path
        description="Example deployment defined with RunnerDeployment",
        parameters={"name": "Prefect"},
        tags=["example"],
        schedules=[create_deployment_schedule_create(Cron("0 * * * *"))],  # hourly
        work_pool_name="my-work-pool",  # replace with your work pool
        # job_variables={"env": {"FOO":"BAR"}}  # optional: pool-specific overrides
    )

    # Optionally set an image (e.g., for Docker/K8s/ECS pools); build/push default True
    [deployment_id] = deploy(d, image="<http://ghcr.io/your-org/your-image:latest|ghcr.io/your-org/your-image:latest>")
    print(f"Created deployment: {deployment_id}")

Prefect 3.x: deployment from a Git repo (source + entrypoint)

from prefect import flow
from prefect.deployments.runner import RunnerDeployment
from prefect.deployments import deploy
from prefect.deployments.schedules import create_deployment_schedule_create
from prefect.schedules import Cron
from prefect.storage import create_storage_from_source

@flow
def etl(name: str = "world"):
    pass  # actual code will be loaded from the repo at run time

if __name__ == "__main__":
    storage = create_storage_from_source("<https://github.com/your-org/your-repo>")

    d = RunnerDeployment(
        name="etl-prod",
        flow_name="etl",
        entrypoint="flows/etl.py:etl",  # path in the repo + ":" + flow name
        storage=storage,
        description="Prod deployment from GitHub",
        parameters={"name": "Production"},
        tags=["prod"],
        schedules=[create_deployment_schedule_create(Cron("0 0 * * *"))],  # daily
        work_pool_name="k8s-pool",
    )

    [deployment_id] = deploy(d, image="<http://ghcr.io/your-org/your-image:latest|ghcr.io/your-org/your-image:latest>")
    print(f"Created deployment: {deployment_id}")

Notes - In Prefect 3.x, the old 2.x

Deployment.build_from_flow()

and

prefect deployment build

were removed. Use

RunnerDeployment

or

flow.deploy(...)

instead. - Make sure a worker is running for your target work pool so runs get picked up. Docs - Deploy via Python: Deploy via Python - Create deployments overview: Create deployments Are you on Prefect 2.x and specifically want the 2.x

Deployment

class pattern? If yes, tell me your storage (e.g., local path, GitHub), schedule, and work queue/pool and I’ll tailor the snippet.