Vector Database

A vector database stores embeddings – numerical representations of content – and enables semantic similarity search: you query not for exact words but for content that is semantically similar. Vector databases are a central component of RAG systems because they quickly find the relevant documents the language model needs for a fact-based answer. Well-known products include Pinecone, Weaviate, or Qdrant.

Related Terms

• Embedding
• Retrieval-Augmented Generation (RAG)
• Large Language Model (LLM)

Sources: a16z AI Glossary

← Back to AI Glossary