# Vector database

> What a vector database is, how AI finds the right passage via embeddings, and why for AI visibility meaning counts rather than keyword density.

Synonyms: embedding database, vector store

A vector database is a store that keeps content not as words but as so-called embeddings – sequences of numbers that represent the meaning of a text. Two passages with similar meaning sit close together in this database, even if they use completely different words.

This is exactly what AI answer machines use: when someone asks a question, the question too is translated into such a numerical pattern and the passage closest in content is retrieved. What matters is therefore not whether a text contains the exact search words, but whether it clearly carries the intended meaning.

For AI visibility this means: clearly worded, topically focused passages are found more readily than ones stuffed with keywords. Meaning beats keyword density.

---
Source: https://btlabs.dev/en/glossary/vector-database
Last-Modified: 2026-08-10T12:22:19.532Z
Languages: [de](https://btlabs.dev/llms/de/glossary/vektordatenbank) · [it](https://btlabs.dev/llms/it/glossary/database-vettoriale)
See also: [llms.txt](https://btlabs.dev/llms.txt) · [ai.txt (Policy)](https://btlabs.dev/ai.txt) · [identity.json](https://btlabs.dev/identity.json)