Document storage,
built for engineers.

A Rust-powered document database designed for backend applications, practical operations, and transparent performance engineering. Official SDK support currently covers JavaScript and TypeScript.

Self-hostable · V2 pre-alpha · APIs and deployment details live in the maintained docs

liorandb / local
$ docker run -d --name liorandb -p 27018:27018 -p 27019:27019 -p 27201:27201 -v ldb-data:/var/lib/liorandb/data liorandb/liorandb:pre-alpha
$ 
$ npm install @liorandb/driver
$ node index.ts

Measured, with context

A benchmark preview you can interrogate.

See the full comparison

Peak raw writes

300K

docs/sec · observed local peak

10M completion

36s

end-to-end insertion time

Read throughput

35K

reads/sec · query shape unrecorded

Memory observed

3.5 GiB

LioranDB only; not a comparison

user-reported development measurement; raw artifacts for the matched comparison are pending publication. These numbers describe one local development workload and should not be read as universal production performance.

Engine surface

A practical storage layer for backend work.

LioranDB keeps the developer path close to the engine: a familiar document API, a Rust storage core, and deployment choices that are explicit about what is ready today.

documented

Document API

Collections, familiar query operators, updates, pagination, and aggregation for application data.

MongoDB-style document model · JavaScript and TypeScript SDK · Collections and indexes
implemented / pre-alpha

Storage engine

A Rust-based engine with WAL and checkpoint paths exposed as part of the V2 architecture direction.

Rust engine · WAL and checkpoints · Persistent local storage
available in test paths

Search and indexing

Secondary and text indexing are available in documented V2 test paths; workload-specific behavior matters.

Secondary indexes · Text index · Compaction and recovery paths
documented

Operations

Run locally with Docker, connect through the dashboard, or review managed deployment options.

Docker quickstart · CLI package · Managed provisioning
version-sensitive

Transport

The official docs describe JavaScript and TypeScript SDK, gRPC, and REST access paths. Verify endpoint details against the docs for your version.

JavaScript / TypeScript SDK · gRPC endpoint · REST endpoint
tested / pre-alpha

Recovery

WAL replay and recovery are part of the documented V2 test surface; durability semantics depend on the configured path.

WAL replay · Crash-recovery test artifacts · Checkpoint lifecycle

Developer path

Start local. Keep the escape hatch.

The first step is a local Docker instance and a documented driver. Managed infrastructure, Studio, and the source repositories are separate paths you can evaluate when the workload earns it.

index.tsCode can be copied to the clipboard.
import { LioranDBClient } from "@liorandb/driver";

const client = await LioranDBClient.connect(
  process.env.LIORANDB_URI ??
    "liorandb://admin:password@127.0.0.1:27018/default",
);

try {
  const users = client.db("default").collection("users");
  await users.insertOne({ email: "ada@example.com", active: true, age: 31 });
  const user = await users.findOne({ email: "ada@example.com" });
  console.log(user);
} finally {
  await client.close();
}

Architecture

Know where the request goes.

Follow the path from client request to persistent state. Implementation status and caveats are called out on the architecture page.

Explore the architecture
01

Client request

JavaScript / TypeScript SDK, REST, or gRPC

02

Document API

Collections, filters, updates, and query execution

03

Storage paths

Primary storage plus configured secondary/text indexes

04

Durability

WAL and checkpoint lifecycle

05

Persistent state

Local or managed deployment storage

Self-host

Run the local Docker path and keep your operational boundary close.

Review deployment

Build with drivers

Use the official JavaScript and TypeScript SDK entry points.

See drivers

Choose managed

Compare shared and dedicated offerings without confusing them with local benchmarks.

Review cloud