is a hosted MLOps product with their own framework. Experiment tracking is in the core.

When to use

When you want to just save machine learning model meta data from other data science platform. pricing pricing can be found here . keeps track of you logging hours. A single logging is 1 second, but any consecutive logging within 10 minutes will charge the actual time interval.

Individual plan is free. 1 member, 200 monitoring hours per month.

Team plan costs 150 $ / month. Unlimited members, 1500 monitoring hours per month.

Organization plan is 600 $ / month. 6000 monitoring hours per month. User management.

Custom is meant for enterprises.


Managed service by default. Capability to deploy to on-premises, AWS, Google or Azure as a Kubernetes cluster.

Parallel computation

Not applicable. The platform does not process data. It has no computation engine for data science work.

Programming languages has client packages for Python and R. review

The company was born after the founders won a Kaggle image recognition competition. is an experiment tracking and MLOps platform. The ML model registry is in the core of the product.

It does not have a built-in Jupyter notebook or similar environment. Rather, you use their API to log the results from whatever data science environment you are using.

Here you find great resources to compare against the competitors.