Databases
Vector search inside a database you already run, and dedicated vector stores.
-
Amazon S3 Vectors
AWS documents S3 Vectors as vector storage in vector buckets and vector indexes, with similarity, full-text, and hybrid queries. AWS states sub-second latency for infrequent queries and as low as 100 milliseconds for more frequent queries.
-
Chroma
Open-source search library (Apache-2.0) and Chroma Cloud. The pricing page describes vector, full-text, and metadata search.
-
ClickHouse vector search
ClickHouse documents exact and approximate vector search for the MergeTree family.
-
Elasticsearch
Elastic's vector search page says Elasticsearch stores embeddings and retrieves similar vectors, and that it can combine that with full-text search, filters, aggregations, and hybrid retrieval.
-
LanceDB
Open-source database (Apache-2.0) built on the Lance format. The public pricing URL redirected to a contact page on the check date.
-
MariaDB Vector
MariaDB documents vector functions in the server SQL reference. This is the in-server option if you already run MariaDB.
-
Milvus
Open-source vector database under Apache-2.0. Zilliz Cloud is the managed service and has its own entry.
-
MongoDB Vector Search
MongoDB documents Vector Search as its retrieval product for AI applications. Confirm on that page which deployments include it.
-
MySQL VECTOR
MySQL documents a VECTOR column type. Each entry is a 4-byte float. The manual says the default length is 2048 and the maximum is 16383.
-
OpenSearch k-NN
OpenSearch publishes a k-NN search plugin for vector search on an OpenSearch cluster.
-
Pinecone
Managed vector database. The docs list full-text, sparse-vector, hybrid search, and filter-by-metadata guides. The pricing page lists Starter, Builder, Standard, and Enterprise.
-
Qdrant
Open-source vector database (Apache-2.0) with a managed cloud. The pricing page's release note for 1.20 mentions BM25 hybrid search without sparse vectors.
-
Redis vector search
Redis documents vector fields and vector search. The page title is vector search concepts and the description says you can perform vector searches in Redis.
-
Typesense
Search engine with nearest-neighbor vector search. The vector search guide lists hybrid search (keyword plus semantic plus filtering) as a use case, and documents a vector_distance from 0 to 2 for cosine.
-
VectorChord
PostgreSQL extension for vector search that depends on pgvector and uses the vector type. The overview documents vchordrq indexes.
-
Vespa
Open-source search engine (Apache-2.0) for ranking and serving. You run the server yourself.
-
VillageSQL Vector
Open-source, self-hosted vector search for VillageSQL, a MySQL-compatible server fork. The vsql-vector-pk extension adds an SVECTOR float32 type (up to 3072 dimensions) and an HNSW index (L2, L1, cosine, and inner product), queried with ORDER BY distance LIMIT k. Experimental / preview: the README says it is not recommended for production use.
-
Weaviate
Vector database you can run yourself or as Weaviate Cloud. Source outside the wl/ directory is BSD-3-Clause. Code under wl/ is the Weaviate enterprise license.
-
Zilliz Cloud
Managed Milvus. The pricing page lists vector search, filtered search, hybrid search, and full text search as data-service features, and publishes a free plan plus dedicated cluster prices.
-
pgvector
Open-source vector similarity search for Postgres. The README documents exact and approximate nearest neighbor search, and says to combine it with Postgres full-text search for hybrid search.
-
pgvectorscale
Postgres extension that adds a StreamingDiskANN index on pgvector data. The README documents label filtering and WHERE post-filtering.
-
turbopuffer
Managed vector and full-text database. The pricing FAQ describes a commercial-only model. The launch plan has a $16/month usage minimum.