What is Chunk Overlap and why use it?

Chunk Overlap is the repeated text shared between adjacent content chunks when documents are split for retrieval. It helps preserve context so a passage does not lose important setup, definitions, or relationships when separated from the surrounding page. In AI search, sensible overlap can improve retrieval quality, especially for technical documentation, long guides, and comparison content. Too little overlap can make chunks confusing; too much overlap can waste tokens and create duplicate retrieval results. Content teams can support better chunking by using descriptive headings, self-contained paragraphs, summaries, and clear transitions that help each section stand on its own.