ParticleDB

159KTPS (TPC-C, 8 workers)

17msClickBench Hot Geomean

25MOps/sec PG Wire

5Wire Protocols

Quick Install

docker run -d --name particledb -p 5432:5432 -p 8080:8080 particledbai/particledb

Connect with any PostgreSQL client:

psql -h localhost -p 5432

Why ParticleDB?

AI Native

Vector search, embeddings, RAG, and classification — built into the SQL engine. No sidecar vector DB, no data sync headaches.

Scalable Transactions

Full ACID with MVCC. 159K TPC-C transactions per second. Concurrent reads never block writes.

Fast Analytics

Columnar storage, zone maps, SIMD aggregation. Sub-millisecond grouped queries on millions of rows.

PostgreSQL Compatible

Drop-in PG wire protocol. Your existing drivers, ORMs, and tools just work — psql, JDBC, SQLAlchemy, Prisma.

Multi-Protocol

PG wire, gRPC, HTTP, and Redis RESP on one engine. Use SSE streaming today for browser-friendly incremental results.

Cloud Ready

Docker, Kubernetes, or bare metal. Deploy to AWS, Azure, or GCP.

SDKs for Every Language

TypeScript, Python, Go, Rust, Java, and C# — first-class support for SQL, vectors, and RAG.

import { ParticleDB } from 'particledb';

const db = new ParticleDB({ host: 'localhost', port: 5432 });

// SQL queries
const users = await db.query('SELECT * FROM users WHERE age > $1', [25]);

// Vector similarity search
const similar = await db.vector.search('embeddings', queryVector, { limit: 10 });

// RAG pipeline
const answer = await db.rag.query('What were Q4 revenue trends?', { table: 'reports' });