Give us a ⭐️ on GitHub to support the project!
Give us a star ⭐️ on GitHub to support the project!
⚡️ Join us June 6 to see how you can use Evidently open-source to evaluate LLMs. Register now →

Evidently Open-Source

An open-source framework to evaluate, test and monitor ML models in production. Prevent incidents, build trust, and make continuous improvements.

GET STARTED

What's in there

Evidently has three components. You can start with minimal effort and one-off checks. As you scale, move to a complete monitoring stack with a self-hosted dashboard.

Evidently Reports

Evidently Reports

Visualize your models and data with rich interactive reports.

Evidently Test Suites

Evidently Test Suites

Run data and ML model checks for production pipelines and CI/CD.

Evidently ML Monitoring

Evidently ML Monitoring

Get a central dashboard to track the health of ML models and datasets.

When to use Evidently

Across the ML model lifecycle. Before deployment, validate your models and data with Reports and Test Suites. In production, gain visibility with Evidently ML Monitoring.

Evidently ML observability platform

What you can evaluate 

Hundreds of checks. From counting nulls in data to detecting embedding drift.

Data quality
Track data quality and integrity.
Learn more →
Data drift
Explore statistical distribution shifts. 
Learn more →
Model performance
Evaluate and compare ML model quality.
Learn more →
NLP and LLM
Monitor text-based models. 
Learn more →
Evidently in-built checks for tabular and text data

Why Evidently

Open-source

  • Your data is yours. You can run Evidently on your infrastructure.
  • No vendor lock-in.
  • No black boxes. You can inspect the entire source code on GitHub.

Fits into any stack

  • Run wherever you run Python.
  • Add to the existing models, workflows, and infrastructure.
  • Batch or real-time. Text or tabular. Cloud or local. Pytorch or Scikit-learn. And anything else!

Start fast, tweak anything

  • Be as hands-off or hands-on as you like.
  • Start with presets and automated thresholds.
  • Customize metrics, dashboards, and test conditions as you go.

Pick metrics, not write them

  • 100+ useful checks and metrics out of the box.
  • In-built visualizations and parameters to tweak.
  • An ever-growing set of metrics, with an option to add yours.

One tool for ML reliability

  • One API to learn and tool to configure.
  • Evaluate drift or accuracy the same way each time.
  • Easily switch from monitoring to debugging.

Component-based

  • Use only what you need.
  • Get going with reports, and switch to complete monitoring setup once ready.
  • Modular architecture. You can extend and build upon it.

Yes, it's open-source!

With a lot of examples in the docs.

GET STARTED

Get started

Easily add Evidently to existing workflows, no matter where you deploy. 

Evidently Cloud

Evidently Cloud is the easiest way to get ML monitoring up and running.

START FREE

Open-source

Deploy and run Evidently on your own.
Apache 2.0 license.

DEPlOY NOW
By clicking “Accept”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.