As this ebook edition doesn't have fixed pagination, the page numbers below are hyperlinked for reference only, based on the printed edition of this book.

While the VectorStoreIndex may be the star of the show in most of our RAG scenarios, LlamaIndex provides many other useful indexing tools. They all have specific features and use cases and the following section will explore them in more detail.

The SummaryIndex

The SummaryIndex offers a straightforward yet powerful way of indexing data for retrieval purposes. Unlike the VectorStoreIndex, which focuses on embeddings within a vector store, the SummaryIndex is based on a simple data structure where nodes are stored in a sequence. You’ll find a simple depiction of the structure of the SummaryIndex in Figure 5.7:

Figure 5.7 – The structure of a SummaryIndex

When building the Index, it ingests a collection of documents, splits them into smaller...