Must fields used in Data Models be extracted before creating the datasets?

The assertion that fields used in Data Models do not need to be extracted prior to creating the datasets is accurate. In Splunk, Data Models are built on top of indexed data and rely on the underlying search-time field extraction capabilities. When creating a Data Model, the fields can be defined dynamically, allowing for the use of fields that are automatically extracted during searches rather than requiring pre-extracted fields.

This flexibility is a significant aspect of how Splunk operates. It allows users to define the fields they want to work with at query time, which can streamline the data modeling process and make it more efficient. Thus, you can leverage both the existing fields in your data and specify new fields as needed without requiring prior extraction.

The other options mentioned imply different requirements for field extraction that do not align with how Splunk Data Models function, making them less applicable in the context of this question.