Skip to content

On-premise users: click in-app to access the full platform documentation for your version of DataRobot.

Capability matrix

The following table provides an evolving comparison of capabilities available in DataRobot Classic and Workbench.

Feature DataRobot Classic Workbench
Generative AI
Add/configure vector databases No Yes
Build/Chat/Compare LLM blueprints No Yes
Evaluation and moderation metrics No Yes
Deploy LLMs No Yes
General platform features
Sharing Data, projects Data, Use Cases
Business-wide solution No, single projects Yes, experiments in a Use Case
Authentication SSO, 2FA, API key management SSO, 2FA, API key management
GPUs (for deep learning) Yes Yes
Data-related capabilities
Data sources Certified JDBC Connectors, local file upload, URL Snowflake, BigQuery, Databricks, S3, local file upload, AI Catalog assets
Data preparation No Wrangling
Feature Discovery Yes Yes
Data Quality Assessment Yes No
Data storage AI Catalog Data Registry
User-created feature lists Yes Yes
Modeling-related capabilities
Modeling types Binary classification, regression, multiclass, multilabel, clustering, anomaly detection Binary classification, multiclass classification, regression
Partitioning Random, Partition Feature, Group, Date/Time, Stratified Random, Stratified, Date/Time, User-defined grouping, Automated grouping
TVH partitioning Yes Yes
Modeling modes Quick, full Autopilot, Comprehensive, Manual Quick, Manual, Comprehensive (predictive only)
Incremental learning No Yes
Advanced options Yes Partitioning, monotonic, weight, insurance-specific
Time-aware Yes, time series and OTV Yes
Blenders Yes, with option enabled No
Retraining Yes New feature list, sample size, training period
Model Repository Yes Yes
Composable ML Yes No
Visual AI Yes No
Bias and Fairness Yes No
Text AI Yes Yes, for supported model types
Location AI Yes No
Model insights See the full list Insights for predictive or time-aware experiments
Prediction Explanations XEMP and SHAP SHAP (predictive only)
Unlocking holdout Automatically for the recommended model or anything prepared for deployment Automatically for all models
Downloads Data, Leaderboard, Scoring Code, Compliance Report, exportable charts Compliance Report
Prediction-related capabilities
Predictions Yes Yes
MLOps Yes Yes
No-Code AI Apps
No-Code AI Apps Yes Yes
DataRobot Notebooks
DataRobot Notebooks Yes Yes
Notebook scheduling No Yes
Codespaces No Yes
Notebook sharing No Yes
DataRobot Generative AI
GenAI No Yes

Updated May 14, 2024