Compare keyword, vector, and hybrid search for RAG.

Combines exact lexical matching with semantic recall and explains why hybrid retrieval often needs measured fusion and ranking.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

Keyword search is strong for exact terms, identifiers, names, and rare phrases, while vector search can find semantically related text with different wording. Hybrid search runs both and combines their ranked results. I choose among them using a query set from the real domain rather than assuming vector search always improves relevance.

Full answer below

Practice focus

Search-method trade-offs, exact identifiers, semantic similarity, and result fusion.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Overview
Keyword and vector retrieval fail differently, which makes their combination useful for many corpora.
Keyword search is strong for exact terms, identifiers, names, and rare phrases, while vector search can find semantically related text with different wording. Hybrid search runs both and combines their ranked results. I choose among them using a query set from the real domain rather than assuming vector search always improves relevance.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
An error code like AUTH-1042 may be best found lexically, while how do I regain access may benefit from semantic retrieval. Hybrid search can recover both, but score scales from different retrievers are not directly comparable. I use a defined fusion method, inspect which retriever contributed each result, and tune weights or candidate counts against relevance judgments.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Using vector search for exact IDs without testing it.
Adding scores from different retrievers directly.
Calling hybrid search better without a relevance dataset.

Compare keyword, vector, and hybrid search for RAG.

Practice focus

Answer path

Overview

Review signals

Follow-up prompts

Continue practice

Next in path

Tech stacks

Skill areas

Human editorial direction

Was this drill useful?

Found this answer useful?