Prediction API¶
DataRobot's Prediction API provides a mechanism for using your model for real-time predictions via serverless or dedicated prediction environments. Follow the guidelines for making serverless predictions with the Prediction API. Serverless predictions use the REST API endpoint /api/v2/deployments/:id/predictions and do not require a datarobot-key header.
To access documentation for the dedicated server Prediction API, navigate to the Dedicated Prediction API page.
| Topic | Description |
|---|---|
| Make predictions with the API (serverless) | Make predictions using serverless prediction environments. |
| Make predictions with the Python API client | Use the DataRobot Prediction Library, a Python library for making predictions with various prediction methods. |
| Get a prediction server ID | Retrieve a prediction server ID using cURL commands from the REST API or the DataRobot Python client. |
| Serverless Prediction API reference | Review Prediction API methods, input and output parameters, and errors for serverless predictions. |
| Make predictions with the API (dedicated server) | Legacy documentation for using the Prediction API with dedicated servers. |
| Deprecated API routes | Review deprecated Prediction API routes in a reference document listing the deprecated requests and their replacements. |