What is Watson Machine Learning?

Watson Machine Learning is a full-service IBM Cloud™ or private cloud offering that makes it easy for developers and data scientists to work together to integrate predictive capabilities with their applications. The machine learning service is a set of REST APIs that you can call from any programming language to develop applications that make smarter decisions, solve tough problems and improve user outcomes.

Watch the video (03:43)

Do the models need to be built using only IBM offerings?

Watson Machine Learning allows the flexibility to import models trained elsewhere and the portability to deploy on the cloud, on premises or on devices. Watson Machine Learning supports models built in the most popular open source ML frameworks, as well as IBM frameworks like IBM SPSS® Modeler and IBM Decision Optimization.

See Watson Machine Learning supported frameworks

Does Watson Machine Learning integrate with Decision Optimization?

Yes! You can now deploy and solve your Decision Optimization models inside the Watson Machine Learning service. For each deployment in the Watson Machine Learning service, you can specify the run capacity you want to allow to this deployment (# VPC, # RAM).

Learn more

What is Watson Machine Learning Local?

Watson Machine Learning Local is a private cloud offering that makes it easy for developers and data scientists to work together to integrate predictive capabilities with their applications. Watson Machine Learning Local is available as an add-on for Watson Studio Local, where it can be integrated to operationalize models and analytical assets.

Learn more

What is Watson Machine Learning Accelerator (formerly PowerAI Enterprise)?

As of February 2019, the Watson Machine Learning family has been expanded with a renamed offering: Watson Machine Learning Accelerator (formerly PowerAI Enterprise). Available as an add-on to the Watson Machine Learning base, Watson Machine Learning Accelerator provides distributed, scale-out processing of machine learning and deep learning workloads in a multitenant, private cloud environment. Together with Watson Studio, the expanded Watson Machine Learning offerings empower businesses to quickly start model management, speed time to deployment and simplify the process of ensuring production-worthy accuracy to help operationalize the data science lifecycle.

Watson Machine Learning Accelerator targets organizations moving from exploration and single-node development to scale-out production environments for machine learning and deep learning technologies. The key technical components include multitenancy, dynamic allocation of compute resources, elastic distributed training, parallel hyperparameter optimization, training visualization, and transparent scaling from a single server to many servers.

Learn more

What is AutoAI?

AutoAI automatically prepares data, applies algorithms and attempts to build model pipelines best suited for your data and use case. AutoAI refers to the automation of AI development, including model building stages and deployment and monitoring of models in production. The tool is included with Watson Studio, and runs training and deployments with Watson Machine Learning.

Learn more

Is Watson Machine Learning available on IBM Cloud Paks?

In 2018, Watson Machine Learning became a part of IBM Cloud Paks™ (formerly ICP and ICP for Data). It is also available on a public or private cloud, or as a fully managed service on IBM Cloud.

Learn more

Who should use Watson Machine Learning?

Data scientists and application developers are the primary users. Watson Machine Learning includes a repository for analytical assets like machine learning models, deployment services for easy deployment and integration of models, and services that assist in the training and monitoring of deployed models. All of these services provide a complete machine learning solution. They can help a new data science team or an advanced team requiring high-throughput deployments and scaling for their model training jobs.

Watch the demo (02:49)

How do I sign up for the free trial?

Signing up for the trial of the Lite plan is easy. Just create a no cost IBM Cloud account here. From there, you will be asked to enter or sign up for an IBM ID. Enter your information, then confirm your email, and you'll have a Watson Machine Learning instance in your account. You can work with this directly using APIs or the Python client library, or log in to Watson Studio to use the user interface to manage models and create deployments. 

Try for free

How long can I use the free trial?

You'll be able to use the Lite plan for as long as you want, with a maximum of five deployed models, 5000 predictions per month, and 50 capacity unit hours per month. The Lite plan is limited based on monthly usage limits that refresh every month.

Try for free

What is the difference between the Standard and Professional plans?

The Standard plan charges a flat rate per each thousand predictions and per total number of capacity unit hours (USD .50/1000 predictions, USD .50/capacity unit-hour). Minimum of 1 minute per training job. 

The Professional plan is billed at a fixed amount per month, including 2,000,000 predictions and 1000 capacity unit hours (USD 1000/instance, USD .40/1000 predictions, USD .40/capacity unit hour). Minimum of 1 minute per training job. Overage is charged at a flat rate per each thousand predictions and per each capacity unit hour.

View pricing

What do I do if I can’t log in? It says my trial expired, but I’ve never used Watson Machine Learning before.

IBM Watson apps are powered by IBM Cloud. You may have previously used an IBM Cloud account that is now expired. To continue your trial, obtain a code by contacting IBM through the chat icon.

How do I continue using Watson apps after my IBM Cloud trial ends?

During your IBM Cloud trial, you can obtain a code to continue your trial by contacting IBM through the chat icon. Alternatively, you can upgrade your IBM Cloud account to the pay-as-you-go plan by providing a credit card number.

Get started on Watson Machine Learning in minutes

Get started at no charge.