Skip to main content
Core concept · AI Search Infrastructure

Definition

Passage indexing is the practice of storing and retrieving individual passages of a page rather than the page as a whole. A search system splits content into passages, indexes each one, and can return a single passage as an answer even when the rest of the page is unrelated. It decides whether a passage exists as a retrievable unit at all. A model answers from a passage, not a URL. Passage indexing is why a single well-formed paragraph can be cited from a long page, and why a strong page with no cleanly separable passages can go unused. It is distinct from passage ranking, which decides the order among passages already retrieved. Indexing decides whether the passage is retrievable in the first place. Structuring content into self-contained passages is what makes it indexable at this grain.

Passage ranking

Distinct from

Chunking

See also

Passage-level optimization

See also

Semantic search

See also

Quotability

See also

Relevant Plate Lunch Collective Services

Citation-Ready Content AI SEO