---
id: azureml-mlflow
version: "1.62.0.post5"
license: MIT
license_treatment: permissive
maintenance: active
---

# azureml-mlflow — Contains the integration code of AzureML with Mlflow.

License: permissive · Maintenance: active · Downloads: 2.7M/mo

## What it is and what it does

azureml-mlflow is a bridge between MLflow and Azure Machine Learning that lets you use MLflow's standard experiment tracking APIs while storing results in an AzureML workspace. It wraps the connection setup so that when you call mlflow.set_tracking_uri() with an AzureML workspace URI, your logged metrics, parameters, and artifacts flow directly into AzureML's tracking backend instead of a local or remote MLflow server.

The package depends on Azure SDK libraries (azure-identity, azure-storage-blob, azure-core) to authenticate and communicate with Azure, plus mlflow-skinny for the core tracking interface. It's designed for teams already using AzureML who want to standardize on MLflow's API without managing a separate MLflow server, or for researchers migrating experiments from local MLflow to cloud-hosted AzureML infrastructure.

Use it for:

- Log experiment metrics and models from a local training script directly to an AzureML workspace using standard MLflow APIs.
- Migrate existing MLflow experiment tracking code to AzureML by changing only the tracking URI configuration.
- Centralize ML experiment tracking across a team by routing all MLflow logs through a shared AzureML workspace.
- Track hyperparameter tuning runs in AzureML while keeping your training code independent of Azure-specific APIs.

## Worth the install?

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

Integrates MLflow experiment tracking and model management with Azure Machine Learning workspaces, allowing you to log metrics and artifacts to AzureML while using MLflow APIs.

Yes, if you use Azure Machine Learning and want to standardize on MLflow's tracking interface. The package is actively maintained, has no known vulnerabilities, and low install friction. It's the official integration layer, so it's the right choice for AzureML+MLflow workflows. Skip it if you're not using AzureML or prefer MLflow's native server setup.

## Install

pip install azureml-mlflow
uv add azureml-mlflow
poetry add azureml-mlflow

## Installing azureml-mlflow

Before you install: Low install friction with a pure-Python wheel. Actively maintained with a release within the last 44 days. Requires 11 runtime dependencies including Azure SDK components and MLflow, which are standard for Azure ML workflows.

License in practice: MIT license permits commercial and private use with minimal restrictions; suitable for most projects without licensing concerns.

Quickstart:

pip install azureml-mlflow

import mlflow
from azureml.core import Workspace

workspace = Workspace.from_config()
mlflow.set_tracking_uri(workspace.get_mlflow_tracking_uri())

Requires an active Azure Machine Learning workspace and valid credentials configured via Workspace.from_config().

Verify before relying:

- Specific MLflow version compatibility constraints beyond what mlflow-skinny provides
- Performance characteristics when logging large artifacts or high-frequency metrics
- Whether all MLflow tracking features are fully supported or if some are limited in Azure context

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 2.7M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

mlflow azure machine learning integration, azureml experiment tracking, mlflow tracking uri azure, azure ml metrics logging, mlflow azureml plugin, azure-integration, ml-experiment-tracking

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