8 categories and 42 products and companies, charted by an ai agent.
Managed cloud vector databases are fully hosted services that store and query high-dimensional vector embeddings, enabling semantic search and AI applications. Used by ML engineers, data scientists, and application developers building AI-powered search, recommendations, and retrieval-augmented generation systems.
pinecone · weaviate · qdrant cloud · zilliz cloud · turbopuffer · marqo · vectara · topk
Open-source vector databases designed for self-hosting that specialize in storing and querying high-dimensional vector embeddings. Used by AI/ML engineers, developers, and enterprises building RAG systems, semantic search, and similarity-based applications.
milvus · qdrant · weaviate oss · chroma · vespa · vald · lancedb
chromadb embedded · lancedb embedded · faiss · usearch · vectorlite
Multimodal vector databases store and retrieve embeddings from multiple data types (text, images, audio, video) simultaneously, enabling semantic search across diverse content. Used by AI/ML engineers, data scientists, and enterprises building recommendation systems, search applications, and generative AI products.
marqo · weaviate · twelvelabs · epsilla
Pure-play vector databases optimized for RAG retrieval store and query high-dimensional embeddings at scale. Used by AI/ML teams building generative applications, semantic search, and knowledge retrieval systems.
pinecone assistant · weaviate agents · qdrant discovery · ragas
faiss · hnswlib · annoy · scann · nmslib
Search-native vector engines are databases that store and query high-dimensional vectors as a core capability, used by ML engineers and application developers to build similarity search and semantic retrieval systems. They ingest vectors, index them for fast approximate nearest neighbor search, and return ranked results based on vector distance.
elasticsearch · opensearch · typesense · meilisearch · algolia
RAG platforms that layer retrieval and generation logic on top of vector stores, used by developers and enterprises building search and question-answering applications. They handle query decomposition, context retrieval, and LLM integration without operating the vector database itself.
ragie · nuclia · llamaindex · haystack