What is Vector store?

A database that stores text as embeddings — numerical vectors — so you can retrieve the most semantically similar items to a query, powering RAG.

How it works

Instead of keyword matching, you embed documents and questions into the same vector space and fetch the nearest neighbors. It's how an agent finds the right context in a large corpus without stuffing it all into the prompt.

Example

In practiceYou embed 10,000 support articles; a user question is embedded and the three closest articles are retrieved and passed to the model.

Vector search or keyword search?

Often both — hybrid retrieval fuses them, catching exact terms and semantic matches. Pure vector search misses rare keywords.

Go deeper: Deploy Your First AI Agent

The complete, step-by-step guide — every command, config, and the reasoning behind it.

Get the guide