MLflow Model Registry is a free model store, API, and UI for managing machine-learning models through their lifecycle. It records versions and connects each to the MLflow run, logged model, or notebook that produced it. Teams can organize models with aliases, tags, and annotations, and transition them through stages such as Staging and Production. For registered models on a remote tracking server, MLflow supports basic HTTP authentication and role-based permissions; basic authentication requires a configured secret key and an admin password of at least 12 characters. With Databricks Unity Catalog, the registry supports centralized governance, access controls, cross-workspace access, and model lineage. An official Helm chart is available for Kubernetes self-hosting. MLflow lists integrations with 100+ tools, including LangChain, OpenAI, and PyTorch, and supports Python, TypeScript/JavaScript, Java, R, and OpenTelemetry. The open-source plan is free forever under the Apache 2.0 license. Artifact storage is not supported.
Who it is for
The registry is described as useful for solo data scientists and large machine-learning platform teams. It suits teams managing model versions, lineage, and lifecycle workflows.
What is good
- Tracks model versions and lineage
- Aliases, tags, and annotations
- Supports Staging and Production transitions
- Free under Apache 2.0 license
- Official Helm chart for Kubernetes
What to know first
- Artifact storage is not supported
- Basic authentication requires configuration and a 12-character password
Verdict
MLflow Model Registry provides versioning, lineage, and lifecycle tools in a free open-source package. Teams that need artifact storage should note that it is not supported.
MLflow Model Registry plans and pricing
All plansCompared on model registry software
- Free plan
- Yes
- Model versioning
- Yes
- Approval workflows
- Yes
- Model lineage
- Yes
- Deployment tracking
- Yes
- Model aliases
- Yes
- Artifact storage
- No
