Prediction Server Authentication

A guide to obtaining credentials for making predictions

Prediction Server Key

When you're ready to make predictions from your AI, you need to authenticate with a prediction server. After training and deploying an AI, you can access the Prediction API.

DataRobot's prediction endpoints use two types of keys:

  • API Key, which is the same key as used in the DataRobot modeling API. Read about it in the API Authentication section.

  • DataRobot key, which is specific to the prediction server.

Note that the DataRobot key is not required in VPC/On-premise DataRobot setups. You only need it in our Managed Cloud environments. This includes Trial, US, and EU cloud environments.

Step 1 - Access the Deployments menu

Select the Deployments tab. You'll see a list of all deployments to which you have access. You will see your deployment in this list.

Step 2: Select Deployment

Click your deployment and then select the Integrations tab. From the list of integrations options, select Scoring Code.

Step 3 - Integrations Menu

In the displayed Integrations Scoring Code page, locate DATAROBOT_KEY and copy the displayed key.
Note that in the same snippet you can also access the API_KEY and DEPLOYMENT_ID.

Step 4: Using the copied key

Pass the keys in request headers:

Authorization: Bearer YOUR_API_KEY
datarobot-key: YOUR_DATAROBOT_KEY

Paste the copied key when deploying your models to make predictions. (You can have a look at the Developer Learning Path in the DataRobot Community for some ideas.)

For language-specific examples you can also refer to our repository of prediction samples in various programming languages on GitHub.

Did this page help you?