Pinecone

Pinecone stores and searches embeddings at scale so teams can add semantic search, recommendations, and retrieval-augmented generation without running their own vector infrastructure.