When PageIndex is used for querying, do we need to load the entire document structure tree? #59
Description
When PageIndex is used for querying, do we need to load the entire document structure tree?
That would consume a lot of tokens, right?
What’s the difference compared to loading the original/raw document directly?