ThunderDB

Get Started GitHub

The Distributed HTAP Database for the AI Era

ThunderDB is a production-grade distributed HTAP database written in Rust that unifies transactional processing, analytical queries, and vector search in a single system. Deploy alongside your existing PostgreSQL, MySQL, MongoDB, or Redis — no rip-and-replace required.

Blazing Fast HTAP

Combine OLTP and OLAP workloads in a single engine with row store, columnar store, and vectorized execution — powered by Rust's zero-cost abstractions.

Multi-Protocol

Connect using PostgreSQL, MySQL, or Redis wire protocols. Use REST, gRPC, GraphQL, or WebSocket APIs. Your existing tools and drivers just work.

AI-Native Vector Search

Built-in HNSW and IVF vector indexes for semantic search, RAG pipelines, and AI/ML embedding storage with native SQL integration.

Zero-Downtime Adoption

Deploy as a companion to existing databases using Change Data Capture (CDC) and Foreign Data Wrappers (FDW). Migrate gradually without disruption.

Distributed by Design

Raft consensus, automatic sharding, multi-region replication, and distributed transactions with two-phase commit — scale horizontally with confidence.

Enterprise Security

TLS encryption, SCRAM-SHA-256 authentication, role-based access control, audit logging, and encryption at rest for production deployments.