Encyclopedia · 311 concepts
Learn this concept as a feed

NLP & Language · intermediate · concept 135 of 311 · Updated Nov 4, 2026

Hybrid Search & Reranking

Hybrid search runs keyword and vector search together and merges the results, because each misses what the other finds. Reranking then has a slower, more accurate model reorder the top candidates. In practice this pairing often improves a RAG system more than changing the language model does, and it is usually the first upgrade worth testing.

Key terms

BM25Vector searchReciprocal rank fusionCross-encoderRecall versus precision

Where you meet it in the real world

Enterprise search, RAG over company documents, e-commerce search