Glossary
Vector database
Definition
A vector database stores embeddings and quickly finds those most similar to a query, using indexes built for nearest-neighbour search.
Why it matters when you’re buying
Vector search sits behind semantic search and RAG. Many existing databases now offer it as a feature, so the choice is often between adding vectors to a database you already run and adopting a dedicated system for scale or speed.
How it relates to other terms
It stores embeddings and serves the retrieval step of RAG.
Where to go next
- Databases: news, companies and community questions
- How to evaluate Databases vendors: criteria, questions to ask and red flags
- RAG and vector search: the latest stories and threads on this topic
Written by Techarda editors · Updated . Think a definition is wrong or missing something? Email engage@techarda.com. All terms.