Teide JS

One Engine. Tables and Graphs.
From Node.js.

Zero-copy Node.js bindings for the Teide columnar engine. Full SQL support, graph pattern matching with SQL/PGQ, vector search, and a fluent TypeScript API — all backed by a native C17 core with no data copying.

Get Started GitHub

Features

01

SQL Engine

Full SQL with aggregations, joins, window functions, subqueries, and CTEs. Execute queries synchronously or asynchronously via executeSync() and execute().

02

Graph Queries (SQL/PGQ)

Create property graphs, match patterns with MATCH, traverse variable-length paths, and find shortest paths — all using the SQL/PGQ standard.

03

Vector Search

Cosine similarity, euclidean distance, and K-nearest-neighbor search with HNSW indexes. Built-in support for embedding-based retrieval.

04

Graph Algorithms

PageRank, connected components, Dijkstra shortest path, Louvain community detection, and clustering coefficients — accessible directly from SQL.

05

Zero-Copy Performance

Native C17 columnar engine exposed through NAPI. Data is shared via TypedArray views directly over the C heap — no serialization, no copying.

06

TypeScript API

Fluent, chainable query builder plus a full SQL interface. Complete type safety with TypeScript definitions for every API surface.

Quick Example

import { Context } from 'teidedb';

const ctx = new Context();

// Load a CSV file into a table
ctx.executeSync("CREATE TABLE t AS SELECT * FROM read_csv('data.csv')");

// Run an aggregation query
const result = ctx.executeSync(`
  SELECT category, SUM(price) AS total
  FROM t
  GROUP BY category
  ORDER BY total DESC
`);

console.log(result.columns, result.nRows);

ctx.destroy();