hera-workflows

What it is and what it does

Hera is a Python SDK that bridges Python code and Argo Workflows, a Kubernetes-native workflow engine. It lets you define containerized workflow templates using Python decorators and functions, then submit them directly to an Argo server running on Kubernetes. Instead of writing YAML manifests, you write Python—functions become reusable script templates, and orchestration logic (task dependencies, parallelism, conditional execution) lives in Python code outside your business logic. The SDK handles the translation to Argo's workflow format and manages submission via HTTP requests to your Argo server.

Typical usage involves decorating Python functions with @script(), composing them into DAGs or other workflow patterns, and calling w.create() to submit. Hera depends on pydantic for validation, requests for HTTP communication with the Argo server, and typing-extensions for runtime type hints. It requires an existing Argo Workflows deployment on a Kubernetes cluster and authentication credentials (typically a Bearer token or port-forwarded access).

Use it for:

• Define multi-step data pipelines in Python and run them on Kubernetes without writing YAML workflow manifests.
• Orchestrate containerized ML training jobs with task dependencies, parallelism, and conditional logic.
• Build GitOps-friendly workflows by generating YAML from Python code for version control and review.
• Automate complex scientific computations or simulations across Kubernetes clusters using Python-native syntax.
• Enable data science teams to express workflow logic in Python while leveraging Kubernetes infrastructure.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Hera is a Python SDK that turns Python functions into Argo Workflows templates and submits them to a Kubernetes cluster, letting you define containerized workflow orchestration in pure Python.

Yes, if you have an Argo Workflows cluster running on Kubernetes and want to define workflows in Python instead of YAML. The package is actively maintained, has no known vulnerabilities, low install friction, and a permissive license. It's production-stable (Development Status 5) and supports current Python versions (3.10–3.14). Install only if you already operate Argo Workflows; it is not a standalone orchestration tool.

Install

hera-workflows on PyPI

pip install hera-workflows

uv add hera-workflows

poetry add hera-workflows

Installing hera-workflows

Before you install

Low friction install with three stable runtime dependencies (pydantic, requests, typing-extensions). Active maintenance with recent releases; last commit 2026-08-09 and 927 repository stars indicate ongoing development.

License in practice

Apache 2.0 permissive license allows free use, modification, and distribution with minimal restrictions; suitable for both open-source and commercial projects.

Quickstart

pip install hera-workflows

from hera.workflows import Workflow, script
from hera.shared import global_config
from hera.auth import ArgoCLITokenGenerator

global_config.host = "http://localhost:2746"
global_config.token = ArgoCLITokenGenerator

@script()
def hello(name: str):
    print(f"Hello {name}")

with Workflow(generate_name="test-", entrypoint="hello") as w:
    hello(name="hello", arguments={"name": "World"})

w.create()

Requires an Argo Workflows server deployed to a Kubernetes cluster and network access to its API endpoint (typically via port-forwarding or authenticated HTTP).

Verify before relying

• Whether the package's optional dependencies (yaml, cli, async-client) are necessary for typical workflows or only for advanced use cases.
• Performance characteristics and scalability limits when orchestrating large numbers of tasks or complex DAGs.

Package facts

Evidence: hera_workflows-7.0.0-py3-none-any.whl

Tags

More Scientific/Engineering packages