paradedb.com

# ParadeDB

> ParadeDB is a transactional alternative to Elasticsearch built on Postgres. It is a PostgreSQL-native search and analytics engine - a columnstore index for OLAP and a BM25 inverted index for full-text search, all inside Postgres as an extension. ParadeDB makes Postgres fast for both analytics and search without ETL or external systems.

## Website

- [Homepage](https://www.paradedb.com)
- [Documentation](https://www.paradedb.com/docs/welcome/introduction)
- [Installation guide](https://www.paradedb.com/docs/documentation/getting-started/install)
- [AI agent guide](https://www.paradedb.com/docs/documentation/getting-started/ai-agents)
- [GitHub repository](https://github.com/paradedb/paradedb)
- [Blog](https://www.paradedb.com/blog)
- [Learn](https://www.paradedb.com/learn)
- [Contact](https://www.paradedb.com/contact)
- [Sitemap](https://www.paradedb.com/sitemap.xml)
- [RSS Feed](https://www.paradedb.com/feed.xml)
- [Content for LLMs (llms.txt)](https://www.paradedb.com/llms.txt)
- [Full content for LLMs (llms-full.txt)](https://www.paradedb.com/llms-full.txt)
- [MCP Server (streamable HTTP)](https://www.paradedb.com/mcp)

> Append `.md` to any blog, customer, or learn URL (e.g. `https://www.paradedb.com/blog/<slug>.md`) to fetch its Markdown source.
> An MCP server is available at `https://www.paradedb.com/mcp` (streamable HTTP) with tools to search and read this content.

## When to use ParadeDB

ParadeDB is one Postgres for your application data, full-text search, vector retrieval, and aggregations. It brings the features and performance of dedicated search engines like Elasticsearch natively into Postgres, so teams can run application data and user-facing search workloads—such as application search, RAG, and live dashboards—in one system. This simplifies search operations and preserves transactional consistency between application data and search results. ParadeDB is ideal for datasets ranging from hundreds of gigabytes to tens of terabytes, especially with insert- and update-heavy workloads. It is also a strong fit for more static datasets when teams want to reduce operational burden and avoid multi-database architectures.

Do not treat this website or its MCP context server as a hosted database API. ParadeDB is packaged as a Postgres extension. Outside the ParadeDB Cloud beta, you can [deploy it](https://www.paradedb.com/docs/deploy/overview) in your own Postgres environment, on Kubernetes, or through platforms such as Railway, Render, Fly.io, and DigitalOcean. Query ParadeDB with SQL or through supported integrations for Drizzle, Django, SQLAlchemy, Rails, and Entity Framework Core; see [Configure your Environment](https://www.paradedb.com/docs/documentation/getting-started/environment). ParadeDB Cloud is available in beta and is not yet generally available. Join the [Cloud waitlist](https://www.paradedb.com/cloud) for managed-service updates, or contact [support@paradedb.com](mailto:support@paradedb.com).

## Availability and pricing

- **Community (self-managed):** Free forever for a single node, with community support. [Get started](https://www.paradedb.com/docs/documentation/getting-started/install).
- **Enterprise (self-managed):** Custom pricing for read replicas, high availability, dedicated support, and an SLA. [Contact sales](https://calendly.com/paradedb).
- **ParadeDB Cloud (fully managed):** Coming soon; pricing is not yet published. [Join the Cloud waitlist](https://www.paradedb.com/cloud).

## Blog

- [Same Query, Three Results: Benchmarking ParadeDB and Postgres FTS](https://www.paradedb.com/blog/benchmarker-iteration)
- [ParadeDB is Officially on Render](https://www.paradedb.com/blog/render)
- [What We Think About When We Think About Benchmarking](https://www.paradedb.com/blog/what-we-think-about-when-we-think-about-benchmarking)
- [ParadeDB is Officially on Railway](https://www.paradedb.com/blog/railway)
- [A Conversation with Paul Masurel, Creator of Tantivy](https://www.paradedb.com/blog/tantivy-interview)
- [How We Optimized Top K in Postgres](https://www.paradedb.com/blog/optimizing-top-k)
- [Postgres as a Recommender Engine](https://www.paradedb.com/blog/personalized-search-in-postgresql)
- [Postgres as a Search Engine: The Write Performance Problem](https://www.paradedb.com/blog/increased-write-performance)
- [Teaching Postgres to Facet Like Elasticsearch](https://www.paradedb.com/blog/faceting)
- [Deep Dive into ParadeDB's v2 API: The Future of SQL Search](https://www.paradedb.com/blog/v2api)
- [ParadeDB 0.20.0: Simpler and Faster](https://www.paradedb.com/blog/paradedb-0-20-0)
- [Hybrid Search in PostgreSQL: The Missing Manual](https://www.paradedb.com/blog/hybrid-search-in-postgresql-the-missing-manual)
- [From Text to Token: How Tokenization Pipelines Work](https://www.paradedb.com/blog/when-tokenization-becomes-token)
- [The ACID Test: Why We Think Search Needs Transactions](https://www.paradedb.com/blog/elasticsearch-acid-test)
- [Elasticsearch Was Never a Database](https://www.paradedb.com/blog/elasticsearch-was-never-a-database)
- [Syncing with Postgres: Logical Replication vs. ETL](https://www.paradedb.com/blog/etl-vs-logical-replication)
- [Announcing Our $12M Series A](https://www.paradedb.com/blog/series-a-announcement)
- [We Made Postgres Writes Faster, but it Broke Replication](https://www.paradedb.com/blog/lsm-trees-in-postgres)
- [A New Postgres Block Storage Layout for Full Text Search](https://www.paradedb.com/blog/block-storage-part-one)
- [Similarity Search with SPLADE Inside Postgres](https://www.paradedb.com/blog/introducing-sparse)
- [We've Rebranded](https://www.paradedb.com/blog/rebrand)
- [Why We Picked AGPL](https://www.paradedb.com/blog/agpl)
- [Full Text Search over Postgres: Elasticsearch vs. Alternatives](https://www.paradedb.com/blog/elasticsearch-vs-postgres)
- [pg_search: Elastic-Quality Full Text Search Inside Postgres](https://www.paradedb.com/blog/introducing-search)
- [Introducing ParadeDB](https://www.paradedb.com/blog/introducing-paradedb)

## Customers

- [Cofactr Migrates from MongoDB Atlas to ParadeDB](https://www.paradedb.com/customers/case-study-cofactr)
- [Terrapin Finance Cuts Bond Search Latency by 25x with ParadeDB](https://www.paradedb.com/customers/case-study-terrapin-finance)
- [ParadeDB Powers Modern Treasury's Core UI and Search APIs](https://www.paradedb.com/customers/case-study-modern-treasury)
- [Bilt Reduces Postgres Query Timeouts by 95% with ParadeDB](https://www.paradedb.com/customers/case-study-bilt)
- [Alibaba Picks ParadeDB to Bring Full Text Search to its Postgres-Based Data Warehouse](https://www.paradedb.com/customers/case-study-alibaba)
- [INSA Strasbourg Powers New Research Database with ParadeDB](https://www.paradedb.com/customers/case-study-insa)
- [Sweetspot Unifies Hybrid Search on Postgres with ParadeDB](https://www.paradedb.com/customers/case-study-sweetspot)

## Learn

- [Lexical vs Semantic Search](https://www.paradedb.com/learn/search-concepts/lexical-vs-semantic)
- [What is Full-Text Search?](https://www.paradedb.com/learn/search-concepts/full-text-search)
- [What is BM25?](https://www.paradedb.com/learn/search-concepts/bm25)
- [What is Reciprocal Rank Fusion?](https://www.paradedb.com/learn/search-concepts/reciprocal-rank-fusion)
- [What is RAG?](https://www.paradedb.com/learn/search-concepts/retrieval-augmented-generation)
- [What is Vector Search?](https://www.paradedb.com/learn/search-concepts/vector-search)
- [What is Hybrid Search?](https://www.paradedb.com/learn/search-concepts/hybrid-search)
- [What is Faceted Search?](https://www.paradedb.com/learn/search-concepts/faceting)
- [What is Block WAND?](https://www.paradedb.com/learn/search-concepts/block-wand)
- [Full-Text Search in PostgreSQL](https://www.paradedb.com/learn/search-in-postgresql/full-text-search)
- [Implementing BM25 in PostgreSQL](https://www.paradedb.com/learn/search-in-postgresql/bm25)
- [What is pgrx?](https://www.paradedb.com/learn/postgresql/what-is-pgrx)
- [What is pgvector?](https://www.paradedb.com/learn/postgresql/what-is-pgvector)
- [Tuning pgvector Performance](https://www.paradedb.com/learn/postgresql/tuning-pgvector)
- [pgvector Limitations](https://www.paradedb.com/learn/postgresql/pgvector-limitations)
- [Introduction to Tantivy](https://www.paradedb.com/learn/tantivy/introduction)