// AI

Embedding

A dense numeric vector representation of text, image or audio, the input format for semantic search and RAG.

In depth

Embedding a corpus is the one-time cost of a RAG system; queries then compute embeddings on the fly and retrieve by cosine similarity. OpenAI, Cohere and Voyage lead the general-purpose embedding market; domain-specific models often outperform.

Related terms

Vector Database

A database that stores and searches high-dimensional embeddings, the retrieval layer of every RAG pipeline.

RAG

Retrieval-Augmented Generation, grounding an LLM's response in retrieved documents to reduce hallucination.

← Back to full A-Z