Lantern

The product is an open-source Postgres vector database and toolkit that adds vector search, BM25 text search, and hybrid search inside Postgres to support AI applications such as RAG and semantic search without requiring a separate vector database. It is for developers and data teams in B2B companies building production AI applications. It is positioned as Postgres extensions with serverless indexing and in-database embedding generation, available as self-hosted open source or as a managed cloud service.

Key features

• Vector search with pgvector
• BM25 text search in Postgres
• Hybrid search with RRF reranking
• Serverless indexing offloaded from database
• In-database embedding generation
• In-database LLM execution
• Support for 20+ embedding models and LLMs
• Automatic vector and LLM column creation
• SQL and ORM integrations
• HNSW indexing with compression support