ZenML
Munich, Germany
What it is
ZenML orchestrates Python-based AI and machine-learning pipelines and agents. It lets teams move pipelines among Kubernetes, Vertex AI, SageMaker, AzureML, or a laptop without rewriting code. ZenML versions step outputs such as models and evaluation datasets, supports rerunning from a selected step, and reuses cached results when code and inputs are unchanged. Its integrations directory lists 66 tools, including LangGraph, OpenAI Agents SDK, PyTorch, MLflow, AWS, and Google Cloud Vertex AI Pipelines. Deployment options include local, SaaS, and self-hosted setups using Docker or Kubernetes Helm. ZenML says it stores metadata while data, artifacts, and compute remain in the customer's infrastructure or VPC. The free Open Source plan includes unlimited executions and projects for self-hosted use. Scale costs $999 a month and includes 2,000 executions, 3 projects, and 5 snapshots. Enterprise pricing is custom, with SSO, custom-role RBAC, audit logs, and air-gapped deployment listed among its controls.
For and against
For
- Runs pipelines across several infrastructure options.
- Versions models and evaluation datasets.
- Reuses unchanged cached step results.
- Offers local, SaaS, and self-hosted deployment.
- Free plan includes unlimited executions and projects.
Against
- Scale costs $999 a month.
- Free plan is self-hosted.
- Managed compute is not offered.
The verdict
ZenML combines pipeline orchestration, artifact versioning, caching, and several deployment options. The free plan is self-hosted, while Scale starts at $999 a month and Enterprise pricing is custom.
Who it suits
The Open Source plan is described for individuals and small teams; Scale is aimed at teams running ML in production. Enterprise is described for organizations at scale.
Plans and prices
Open Source
Unlimited executions · Unlimited projects · Self-hosted
Read 30 Sep 2026 · zenml.io
Scale
2,000 executions · 3 projects · 5 snapshots
Read 30 Sep 2026 · zenml.io
Enterprise
Unlimited executions · Unlimited projects · SSO · Custom-role RBAC
Read 30 Sep 2026 · zenml.io
On the record
- Artifact versioning
- ZenML versions step results, including models and evaluation datasets, and supports re-execution from any step.zenml.io · read 30 Sep 2026
- Caching
- ZenML caches step results and reuses them when code and inputs have not changed.zenml.io · read 30 Sep 2026
- Data handling
- ZenML says it stores metadata while data, artifacts, and compute stay in the customer's infrastructure or VPC.zenml.io · read 30 Sep 2026
- Deployment
- ZenML supports local, SaaS, and self-hosted deployment, including Docker and Kubernetes Helm deployment methods.zenml.io · read 30 Sep 2026
- Enterprise controls
- The Enterprise plan adds SSO, custom-role RBAC, audit logs, and air-gapped deployment.zenml.io · read 30 Sep 2026
- Infrastructure
- Pipelines can run on Kubernetes, Vertex AI, SageMaker, AzureML, or a laptop without rewriting the code.zenml.io · read 30 Sep 2026
- Integrations
- The integrations directory lists 66 tools, including LangGraph, OpenAI Agents SDK, PyTorch, MLflow, AWS, and Google Cloud Vertex AI Pipelines.zenml.io · read 30 Sep 2026
- Overage
- ZenML says it does not cut off service for occasional monthly execution overages and may contact customers if overages become a pattern.zenml.io · read 30 Sep 2026
- Purpose
- ZenML orchestrates AI and machine-learning pipelines and agents written in Python.zenml.io · read 30 Sep 2026
- Security
- ZenML states that it is SOC 2 Type II and ISO 27001 certified.zenml.io · read 30 Sep 2026
- Support
- The Open Source plan includes community support, Scale includes priority support, and Enterprise includes dedicated support with an SLA.zenml.io · read 30 Sep 2026
- Target users
- The Open Source plan is described for individuals and small teams, Scale for teams running ML in production, and Enterprise for organizations at scale.zenml.io · read 30 Sep 2026
