FalkorDBLite for TypeScript

FalkorDBLite for Node.js/TypeScript launches an embedded redis-server with the FalkorDB module and returns a connected FalkorDB client. It is ideal for local development, demos, CI jobs, or small apps that want a zero-config graph database that starts and stops with the application.

Requirements

  • Node.js 20 or later
  • Ability to download the FalkorDB module during npm install
  • Linux x64 and macOS arm64 are supported; Windows users should run under WSL2 or use a remote server

Installation 

npm install falkordblite

# Optional: also install the remote client when you plan to connect to an external server
npm install falkordb

Quick start 

import { FalkorDB } from 'falkordblite';

const db = await FalkorDB.open();
const graph = db.selectGraph('quickstart');

await graph.query('CREATE (p:Person {name: "Ada"})');
const result = await graph.roQuery('MATCH (p:Person) RETURN p.name');
console.log(result.data); // => [ [ 'Ada' ] ]

await db.close();

Persist data between runs

Provide a path to keep data on disk. When set, FalkorDBLite enables periodic snapshots automatically.

const db = await FalkorDB.open({ path: '/tmp/falkordb-lite' });
const graph = db.selectGraph('inventory');
await graph.query('CREATE (:Product {id: 1, name: \"Laptop\"})');
await db.close();

Work with multiple graphs

Use separate graph IDs to isolate datasets within the same embedded instance.

const db = await FalkorDB.open();

const users = db.selectGraph('users');
const orders = db.selectGraph('orders');

await users.query('CREATE (:User {id: 1, email: \"a@example.com\"})');
await orders.query('CREATE (:Order {id: 10, total: 99.5})');

await db.close();

Configuration options

Set options on FalkorDB.open() to control the embedded server:

Option Type Default Purpose
path string temporary directory Data directory; enables persistence and snapshots.
redisServerPath string auto Use a custom redis-server binary.
modulePath string auto Use a custom FalkorDB module (.so) path.
maxMemory string unset Redis maxmemory, e.g. "256mb".
logLevel 'debug' \| 'verbose' \| 'notice' \| 'warning' unset Redis log level.
logFile string standard output Where the embedded server logs.
timeout number 10000 Startup timeout in milliseconds.
additionalConfig Record<string, string> none Extra redis.conf entries (e.g. { port: '6379' }).
falkordbVersion string v4.16.3 FalkorDB module release tag to download.
inheritStdio boolean false Pipe redis-server output to the parent process.

The FalkorDB client returned by selectGraph() exposes the full FalkorDB Graph API (query, roQuery, indexes, constraints, profiling, etc.).

Migrate to a remote FalkorDB server

Keep your graph logic and swap only the connection line:

// Embedded
import { FalkorDB } from 'falkordblite';
const db = await FalkorDB.open();

// Remote
import { FalkorDB as RemoteFalkorDB } from 'falkordb';
const remote = await RemoteFalkorDB.connect({
  socket: { host: '127.0.0.1', port: 6379 },
});

Resources

Frequently Asked Questions 5
What Node.js version does FalkorDBLite for TypeScript require?

FalkorDBLite for TypeScript requires Node.js 20 or later. It supports Linux x64 and macOS arm64. Windows users should run under WSL2.

How do I enable data persistence in FalkorDBLite TypeScript?

Pass a path option to FalkorDB.open(): const db = await FalkorDB.open({ path: './my-data' }). This enables periodic snapshots and data persists between application restarts.

Can I control memory usage of the embedded server?

Yes. Use the maxMemory option: FalkorDB.open({ maxMemory: '256mb' }). This sets the Redis maxmemory configuration for the embedded instance.

How do I migrate from FalkorDBLite to a remote FalkorDB server?

Swap the import and connection line. Change from import { FalkorDB } from 'falkordblite' with FalkorDB.open() to import { FalkorDB } from 'falkordb' with FalkorDB.connect({ socket: { host, port } }).

What happens if the FalkorDB module fails to download during npm install?

Check your network connectivity and ensure the target platform is supported (Linux x64 or macOS arm64). You can also provide a custom module path via the modulePath option.