ParadeDB

ParadeDB is a modern alternative to Elasticsearch, built as a Postgres extension that brings Elastic-quality search capabilities directly into PostgreSQL. It supports full-text, hybrid, vector, and faceted search without requiring a separate search infrastructure. ParadeDB can run as a logical replica of any managed Postgres (RDS, Supabase, Neon, Google Cloud, Azure) or be installed inside any self-hosted Postgres deployment with zero additional infrastructure overhead.

Key Features:

• BM25 Full-Text Search — Create BM25 indexes using standard SQL syntax to enable fast, relevance-ranked full-text search natively in Postgres.
• Hybrid Search — Combine lexical (BM25) and semantic (vector) search using Reciprocal Rank Fusion for comprehensive, high-quality results.
• Faceted Search & Aggregations — Build faceted navigation and aggregation queries directly in Postgres without external tooling.
• Advanced Tokenization — Choose from 12+ tokenizers and support for 20+ languages, including dictionary-based and stemming tokenizers.
• Zero ETL Logical Replication — Replicate from any managed Postgres instance to ParadeDB with no ETL pipelines or data duplication overhead.
• Boolean Queries & Top-N Retrieval — Execute complex boolean search queries and efficiently retrieve top-N results using Block-Max WAND dynamic pruning.
• Postgres Extension Install — Install ParadeDB as a Postgres extension into any self-managed Postgres cluster with a single command.
• Docker Deployment — Spin up ParadeDB instantly using the official Docker image for quick testing and development.
• Open Source — ParadeDB is fully open source with 8K+ GitHub stars and 500K+ Docker deployments, backed by an active Slack community.
• Enterprise Features — High availability, read replica support, dedicated SLA, and a fully managed cloud option (private beta) for production workloads.