AI Glossary
Vector database
vector database, vector store
A vector database is a system that stores embeddings and quickly finds the vectors closest to a query by semantic similarity — the foundation of semantic search and RAG systems.
- It stores data as vectors, not as text.
- It searches by semantic similarity, not by an exact word match.
- It serves as the memory layer for RAG and semantic search.
A vector database solves a problem that a classic database doesn't handle well: how to find passages that are similar in meaning rather than identical word for word. Instead of comparing texts, it compares numerical vectors and returns those closest to the query.
In practice, documents are first turned into embeddings and then stored in the database together with metadata. When a question comes in, the system computes similarity and returns the few most relevant passages, which then go to the model as context in a RAG architecture.
Related terms
In guides
Related articles