Vespa

Vespa is an open-source platform for building large-scale AI-powered search and recommendation systems. Originally developed at Yahoo, it combines vector search, full-text search, and structured filtering with machine-learned ranking and inference, all in a single engine that serves queries with low latency over billions of documents. Vespa Cloud offers a managed version for teams that do not want to operate clusters themselves.

Vespa is chosen by organizations with demanding search, recommendation, and RAG workloads that outgrow simpler vector databases. It supports real-time indexing, tensor computation for ranking, and complex queries mixing semantic and keyword signals. Its ability to run inference inside the serving layer makes it powerful for personalized ranking and hybrid retrieval at scale.