ZenML

Product overview

ZenML is an open-source MLOps framework that lets teams define machine learning pipelines as Python code and run them on any infrastructure, from local machines to Kubernetes clusters and cloud environments. The framework acts as a control plane between ML code and the underlying stack, making pipelines portable across orchestrators, experiment trackers, model registries, and deployment targets. The open-source version (Apache 2.0) is free and provides core pipeline orchestration, artifact versioning, and a basic dashboard. Managed SaaS plans add a visual Model Control Plane, advanced scheduling, role-based access controls, and remote development environments. A Pro Self-Hosted tier supports air-gapped deployments with SSO and custom SLAs. The managed SaaS starts at $399 per month for the Starter tier. ZenML supports LLM fine-tuning workflows, RAG pipeline orchestration, and integration with agent frameworks such as LangGraph and CrewAI. The platform connects to experiment trackers, model deployers, and registries across the ML ecosystem, avoiding lock-in to any single vendor. Founded in Munich in 2021 by Adam Probst and Hamza Tahir, ZenML raised $6.4 million from Crane Venture Partners, Point Nine, and angel investors including Richard Socher and Pieter Abbeel. The company offers special pricing for startups and academic institutions. KEY FEATURES: - Open-source MLOps framework (Apache 2.0) with Python-native pipeline definitions - Portable pipelines across Kubernetes, Slurm, and cloud orchestrators - Model Control Plane with artifact versioning and metadata tracking - LLM fine-tuning and RAG pipeline support - Air-gapped and self-hosted deployment options

Sovereignty