What is reranking?
Answer out loud or in your head first, then compare.
Answer
The first search picks a group of possible answers. Reranking is a more careful assessment of those candidates and putting them in a better order. You can use a model that compares the question with each passage for this.
Example: from a hundred similar documents we first pick twenty, then judge more carefully which really answer the question. This can improve quality, but it adds time and cost. It won't fix a document the first stage didn't find at all. Hybrid search is a related but different concept: it combines, for example, keyword search results with vector comparison results.
Why it matters: Explains two frequently used terms without treating them as the same.
Possible follow-ups: Why isn't it worth always reranking every document?
How did it go? Be honest — “Not yet” puts the question on your review list.
Further reading
- Azure AI Search: Hybrid search overview(opens in a new tab)Docs; accessed 22.10.2026. Vectors plus full text, with rankings merged by RRF.