Jump to

Vertex AI Workbench

The single development environment for the entire data science workflow. 

  • Natively analyze your data with a reduction in context switching between services

  • Data to training at scale. Build and train models 5X faster, compared to traditional notebooks

  • Scale up model development with simple connectivity to Vertex AI services


Easy exploration and analysis

Simplified access to data and in-notebook access to machine learning with BigQuery, Dataproc, Spark, and Vertex AI integration.

Rapid prototyping and model development

Take advantage of the power of infinite compute with Vertex AI training for experimentation and prototyping, to go from data to training at scale.

End-to-end notebook workflows

Using Vertex AI Workbench you can implement your training, and deployment workflows on Vertex AI from one place.

Key features

Key features

Fully managed compute

A Jupyter-based fully managed, scalable, enterprise-ready compute infrastructure with security controls and user management capabilities.

Interactive data and ML experience

Explore data and train ML models with easy connections to Google Cloud's big data solutions.

Portal to complete end-to-end ML training

Develop and deploy AI solutions on Vertex AI with minimal transition.

What's new

The latest news and events for Vertex AI Workbench


Technical resources

Google Cloud Basics
Vertex AI Workbench documentation

Learn more about Vertex AI Workbench. 

Google Cloud Basics
Vertex AI documentation

Explore Vertex AI product documentation, from introductory to advanced.

Google Cloud Basics
Explore end-to-end ML on Vertex AI Workbench in this Codelab

In this lab, you'll learn how to use Vertex AI Workbench for data exploration and ML model training.

Google Cloud Basics
Build an image classification model in this Codelab

In this lab, you'll learn how to configure and launch notebook executions with Vertex AI Workbench. 

Google Cloud Basics
Overview of Vertex AI Workbench at Google Cloud Next 2021

Watch "Vertex AI is now even easier for developers" from Google Cloud Next 2021 to explore how Vertex AI Workbench is helping customers to deploy more useful models.

Google Cloud Basics
Create a managed notebooks instance

This in-console tutorial takes you through a step-by-step guide about how to create a managed notebooks instance.

Google Cloud Basics
Create a user-managed notebooks instance

This in-console tutorial takes you through a step-by-step guide about how to create a user-managed notebooks instance.

All features

All features

Simplified data access Extensions will seamlessly connect to the entire data estate including BigQuery, Data Lake, Dataproc, and Spark. Seamlessly scale up or scale out depending on your analytic and AI needs.
Explore data sources using a catalog Write SQL, Spark queries from a syntax-aware, auto-complete enabled notebook cell.
Data visualization Integrated, intelligent visualization tools will provide easy insights into data. 
Hands-off, cost-effective infrastructure All aspects of the compute are managed. Idle timeout and auto shutdown will optimize total cost of ownership.
Enterprise security, simplified Out-of-the-box Google Cloud security controls. Single sign-on and simple authentication to other Google Cloud services.
Data Lake and Spark in one place Whether you use TensorFlow, PyTorch, or Spark, you can run any engine from Vertex AI Workbench. 
Deep Git, training, and MLOps integration With few clicks, plug notebooks into established Ops workflows. Use notebooks for distributed training, hyper-parameter optimization, or scheduled or triggered continuous training. Deep integration with Vertex AI services brings MLOps into the notebook without the need to rewrite code or new workflows.
Seamless CI/CD Kubeflow Pipelines integration to use Notebooks as an ideal, tested, and verified deployment target. 
Notebook viewer Share output of periodically updated notebook cells for reporting and bookkeeping purposes.



Vertex AI Workbench's pricing details can be found here.

The pricing model is based upon compute and services based on the infrastructure you use, as well as other services consumed from within Vertex AI Workbench.