What Is Semantic Search?

Semantic search retrieves by meaning rather than by matching words. Both the query and the documents are turned into embeddings, and the system returns the documents whose vectors sit closest to the query’s — so "how do I reset my password" finds a doc titled "account recovery steps" that shares no keywords. It is the retrieval half of modern question-answering, and the reason an agent can find a relevant passage it could never have located by exact-match search.

Matches on
Meaning (vector proximity), not words
Needs
Embeddings and a vector index
Beats keywords at
Paraphrase and synonymy

Where it wins and where it does not

Keyword (lexical) search excels at exact terms — a part number, a name, a rare token — and fails when the query and the answer use different words for the same thing. Semantic search is the inverse: strong on paraphrase and intent, weaker on exact identifiers it was never trained to distinguish. The dense-retrieval line of work established that learned embeddings beat classic lexical scoring on open-domain question answering.

Because each is strong where the other is weak, serious systems run hybrid retrieval — combining lexical and semantic scores — rather than treating semantic search as a replacement. Presenting it as strictly superior to keyword search is a common and costly oversimplification.

Related standards

Karpukhin et al., 2020 — Dense Passage Retrieval

Questions

Does semantic search replace keyword search?

No — they are complementary. Hybrid retrieval combining both usually beats either alone, especially where exact identifiers matter.

Is semantic search the same as RAG?

No — semantic search is the retrieval step; RAG is the full pattern of retrieving and then generating an answer from what was retrieved.

Keep reading

related
What Is an Embedding?
related
What Is a Vector Database?
related
Retrieval-Augmented Generation (RAG) for Agents
related
RAG vs Long Context: How Should a Model Get Its Knowledge?
referenced by
What Is Tokenization in AI (Subword Tokens)?
Agent Data & Memory
What Is Agent Memory?
Agent Data & Memory
What Is Agent Context (and the Context Window)?
Agent Data & Memory
What Is an Agent Knowledge Graph?
Agent Data & Memory
Managing an Agent’s Context Window
Agent Data & Memory
What Is a Mixture-of-Experts Model?
Agent Data & Memory
What Is Content Addressing?
Agent Data & Memory
What Is Fine-Tuning?
Agent Data & Memory
What Is an AI Hallucination?
Agent Data & Memory
RAG vs Fine-Tuning: Which Should You Use?
Agent Data & Memory
What Is a Transformer (Neural Network Architecture)?
Agent Data & Memory
What Is the Attention Mechanism?
Agent Data & Memory
What Is Chain-of-Thought Prompting?
Agent Data & Memory
What Is RLHF (Reinforcement Learning from Human Feedback)?
Agent Data & Memory
What Is Model Distillation?
Agent Data & Memory
What Is Model Quantization?
Agent Data & Memory
Transformer vs RNN: What Changed?
Agent Data & Memory
Supervised vs Unsupervised Learning: What’s the Difference?

By Michael Gord · published 2026-10-09 · part of the Agentic Encyclopedia. Dates are the day of publication; events are cited at their own dates.