All Tools
P
Vector DBFreeOpen Source
PARADEDB
PostgreSQL for search and vector retrieval with native extensions
AGPL-3.0
ABOUT
Building search and RAG applications typically requires stitching together separate systems: PostgreSQL for application data, Elasticsearch for full-text search, and Pinecone/Weaviate for vector retrieval — each with its own query language, deployment, and operational complexity. ParadeDB extends PostgreSQL with pg_search for BM25 full-text indexing and pgvector-compatible HNSW vector search, enabling developers to run hybrid search (keyword + semantic) alongside their application data in a single Postgres database with standard SQL queries.
INSTALL
docker compose upINTEGRATION GUIDE
1. Run hybrid search combining full-text BM25 scoring with vector similarity in a single PostgreSQL query
2. Replace Elasticsearch with native PostgreSQL full-text search using ParadeDB's pg_search extension
3. Build RAG applications with HNSW vector indexes stored alongside relational application data in Postgres
4. Add semantic search capabilities to existing PostgreSQL applications without introducing a separate vector database
TAGS
postgressearchvector-searchfull-text-searchbm25hnsw