How Documents Are Processed

Once a document is uploaded or saved, it doesn't stay as one long block of text. It's automatically broken into smaller, overlapping segments — roughly 2,000 characters each — before it's made searchable. These segments are what you'll see later as chunks  in the chunk viewer.

The overlap between segments means a passage that falls near a boundary isn't cut off from its surrounding context in every chunk. You don't need to do anything to make this happen — it runs automatically as soon as a document is added, for both uploaded files and text documents you type in directly.

See Viewing & Deleting Chunks  to look at the chunks a document produced.