---
id: kfp-pipeline-spec
version: "2.17.0"
license: Apache 2.0
license_treatment: permissive
maintenance: active
---

# kfp-pipeline-spec — Kubeflow Pipelines pipeline spec

License: permissive · Maintenance: active · Downloads: 4.8M/mo

## What it is and what it does

kfp-pipeline-spec is a low-level component of the Kubeflow Pipelines ecosystem that defines and serializes machine learning pipeline specifications using Protocol Buffers. It provides the data structures and protobuf definitions that represent a Kubeflow pipeline's configuration—its tasks, dependencies, parameters, and execution metadata—in a format that can be validated, stored, and transmitted across the Kubeflow control plane and Kubernetes cluster.

This package is primarily used internally by the Kubeflow Pipelines SDK and platform components rather than directly by end users. It enables the pipeline specification layer that sits between high-level pipeline definitions (written in Python) and the underlying Kubernetes and Argo Workflows execution engines. If you are building pipelines with Kubeflow, you will interact with this package indirectly through the main SDK; direct use is typically limited to advanced scenarios like custom pipeline compilation or specification inspection.

Use it for:

- Serialize and validate Kubeflow pipeline definitions before submission to a Kubeflow cluster.
- Inspect or programmatically manipulate pipeline specifications in protobuf format.
- Build custom pipeline compilation or transformation tools on top of Kubeflow's specification layer.
- Integrate Kubeflow pipeline specs with external ML workflow systems or metadata stores.
- Develop Kubeflow extensions or plugins that need to work with pipeline specifications.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides the pipeline specification and protobuf definitions for Kubeflow Pipelines, enabling serialization and validation of ML workflow configurations on Kubernetes.

Yes, if you are building or extending Kubeflow Pipelines infrastructure or need to work directly with pipeline specifications. No, if you are only writing and running standard ML pipelines—install the main Kubeflow Pipelines SDK instead, which includes this as a dependency. The package is actively maintained, has no known vulnerabilities, and carries permissive licensing; the decision hinges on whether your use case requires direct access to the specification layer.

## Install

pip install kfp-pipeline-spec
uv add kfp-pipeline-spec
poetry add kfp-pipeline-spec

## Installing kfp-pipeline-spec

Before you install: Low installation friction with a single runtime dependency (protobuf). Active maintenance with recent releases; last commit 2026-08-14 and latest release 2026-07-09 indicate ongoing development.

License in practice: Licensed under Apache 2.0 (permissive), allowing use in commercial and proprietary projects with minimal restrictions beyond attribution and license inclusion.

Quickstart:

pip install kfp-pipeline-spec

from kfp_pipeline_spec import pipeline_spec_pb2

# Define or load a pipeline specification
spec = pipeline_spec_pb2.PipelineSpec()

Requires Python 3.9.0 or later.

Verify before relying:

- Whether this package is typically used standalone or always as part of the full Kubeflow Pipelines SDK installation.
- What specific pipeline specification versions or features are supported in 2.17.0.
- Whether protobuf version constraints exist beyond what pip resolves automatically.

## Package facts

- License: Apache 2.0 (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 4.8M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

kubeflow pipeline spec, ML workflow orchestration kubernetes, pipeline specification protobuf, kubeflow pipelines SDK, machine learning pipeline definition, kubernetes ML orchestration, pipeline serialization format, kubernetes, ml-orchestration, protobuf

[View on SkillFed](https://skillfed.io/packages/kfp-pipeline-spec) · [View on PyPI](https://pypi.org/project/kfp-pipeline-spec/)