What is a vector database?
A vector database stores embeddings and quickly finds the ones closest to a query embedding. It's the search engine behind most RAG systems: given a question, it returns the passages whose meaning is most similar, even across millions of documents.
How does a vector database work?
Each document chunk is stored with its embedding. At query time, the database runs a nearest-neighbour search, using clever indexes so it doesn't have to compare against every single item.
Why does a vector database matter?
It makes meaning-based search fast enough for real apps, so AI can pull in the right context in milliseconds.
An example
Example: A legal-tech startup stores every judgment's embeddings in a vector database so lawyers can find similar cases by describing them.
Related terms
Learn it properly
A vector database is covered in AI Builder, with short lessons, practice questions and instant feedback.