Skip to content
GlossaryConceptUpdated May 2026

Vector Store

noun · also: rag, embedding, openai

What is vector store?

A vector store is a database optimized for storing and searching by similarity rather than exact match — the foundation of semantic search and RAG.

Definition

Full definition of vector store

Vector stores hold embeddings (numerical representations of text) and let you query "find the 10 most similar to this query". Pinecone, Weaviate, Qdrant, pgvector are the popular options. Tiny Command integrates with the major ones. Useful for semantic search, RAG, recommendation engines, or any "find related" workflow.

In practice

Vector Store examples

Vector query
Find top 5 documents similar to 'how do I refund a customer?' from the help-articles index.
Used by

Apps that exemplify vector store

See vector store in action across real integrations.

FAQ

Common questions about vector store

Which vector store should I pick?
Pinecone for managed-cloud + easy setup. Weaviate for open-source / self-hosting. pgvector if you already use Postgres and want one fewer service.
How big can the vectors be?
OpenAI's embeddings are 1536 or 3072 dims. Most stores handle millions of vectors per index without trouble.