Vector database
Also: vector store
A database that stores embeddings and quickly finds the ones closest to a query. It is the retrieval half of a RAG system.
A vector database stores embeddings, the numerical representations of your text, and specialises in one operation: given a query embedding, find the stored ones most similar to it. That similarity search is how a RAG system retrieves the passages most relevant to a question.
Options range from libraries embedded in your application, such as Chroma, LanceDB, or sqlite-vec, to standalone servers like Qdrant, and the pgvector extension for Postgres. For local, personal-scale RAG an embedded store is often enough; larger or multi-user systems tend towards a dedicated server.