AxioDB
The Embedded Database for Node.js
Replaces SQLite, LowDB, NeDB & raw JSON files with a real database. MongoDB-style queries, ACID transactions, zero native dependencies. No more node-gyp failures. No more electron-rebuild. Just npm install. Runs on Node.js 20+ and Bun— one codebase, same data files, no runtime lock-in.
All 12 engine test suites — CRUD, transactions, reads, indexed queries & searches, auth, TCP/TLS, HTTP API, and crash recovery — pass on the machine's Node (v26.8.1) and Bun (v1.4.0), including the worker-thread data paths (reads of ≥100-file collections, searches over ≥10,000 documents). Node 20+ is the supported baseline; Bun is verified on the installed v1.4.0 only. Deno passes 9/12 — worker threads are pending there.
Load the AxioDB Agent Skill into ChatGPT, Claude, Cursor, Copilot, or any AI coding assistant to get expert guidance on queries, schema design, transactions, and best practices.
Spin up AxioDB on a cloud container and let your AI agent (Claude, or any MCP-compatible client) talk to that database directly — 43 tools, real login, the exact same RBAC as the web GUI.
CLI tool for AxioDB — interactive REPL with MongoDB shell syntax, all 32 TCP commands, TLS support, and installers for 12 platforms.
Native desktop app (Electron) with a card-based document viewer, bouncy-ball splash loader, connection management, and all-in-one database tooling for Linux, macOS, and Windows.
Insert Operations
Insert Single
InsertMany (500)
Validation
Read/Query Operations (100K docs)
Indexed
documentId
findOne
$gt
$in (5)
Regex
Update Operations
UpdateOne
UpdateMany
Verify
Delete Operations
DeleteOne
DeleteMany
Verify
Transaction Operations
TX Insert
TX Update
TX Delete
TX Mixed
Rollback
Savepoint
withTX
Index Sync
Test Suite Overview (100K docs - 13/13 Passing)
CRUD
3.7 s30 tests
Transactions
625 ms22 tests
Read / Query
146.8 s40 tests
Aggregation
458 ms50 tests
Auth & RBAC
6.0 s32 tests
HTTP API
1.5 s38 tests
TCP Auth
2.7 s20 tests
TCP No-Auth
627 ms6 tests
TCP TX
1.4 s17 tests
TCP TLS
641 ms3 tests
Crash Recovery
5.7 s3 tests
MCP Confirm
40 ms11 tests
MCP Functional
1.7 s6 tests
Hello World with AxioDB
1// npm install axiodb2const { AxioDB } = require('axiodb');3
4// Create AxioDB instance with built-in GUI5const db = new AxioDB({ GUI: true }); // Enable GUI at localhost:270186
7// Create database and collection8const myDB = await db.createDB('HelloWorldDB');9const collection = await myDB.createCollection('greetings');10
11// Insert and retrieve data - Hello World!12await collection.insert({ message: 'Hello, Developer!' });13const result = await collection.query({}).exec();14console.log(result.data.documents[0].message); // Hello, Developer!Introducing AxioDBCloud
Deploy AxioDB in Docker or Cloud. Connect from anywhere with TCP protocol. Same API, zero code changes!
Database Visualization Built In
Start AxioDB with new AxioDB({ GUI: true }) to enable the built-in web GUI on localhost:27018. Perfect for Electron apps—give your users a database inspector without extra dependencies.
Born from SQLite Pain. Built to Be Better.
It started with a simple error: node-gyp ERR! — again.
We were building an Electron app. Needed a database. Tried better-sqlite3 — node-gyp compilation failed. Tried LowDB — data corrupted on concurrent writes. Tried NeDB — abandoned since 2016.
So we built AxioDB.
No node-gyp, no electron-rebuild, no platform headaches
Crash recovery, savepoints, write-ahead logging
JavaScript objects, not SQL strings
InMemoryCache + web dashboard included
Still Using LowDB, NeDB, or better-sqlite3?
Here's what developers actually deal with every day:
- ✗No concurrency — concurrent writes corrupt data
- ✗No ACID transactions — crash = data loss
- ✗No built-in caching — full file rewrite every time
- ✗Single JSON file — entire DB loaded into memory
- ✗Last updated 3 years ago
- ✗No longer maintained since 2016
- ✗Data loss issues reported by users
- ✗File corruption on concurrent access
- ✗No TypeScript support
- ✗Security vulnerabilities unpatched
- ✗Requires native C bindings compilation
- ✗node-gyp headaches on every platform
- ✗electron-rebuild required for Electron apps
- ✗SQL strings instead of JavaScript objects
- ✗Platform-specific builds (Windows ≠ Mac)
AxioDB solves all of these problems:
Why AxioDB?
SQLite requires native C bindings that cause deployment headaches. JSON files have no querying or caching. MongoDB needs a separate server. AxioDB combines the best of all: embedded like SQLite, NoSQL queries like MongoDB, intelligent caching built-in.
Zero native dependencies. No compilation, no platform-specific binaries, no node-gyp headaches. Works everywhere Node.js runs.
Built-in InMemoryCache with automatic invalidation. Instant query results for frequently-accessed data. Multi-core parallelism with Worker Threads.
JavaScript objects, not SQL strings. Operators like $gt, $regex, aggregation pipelines, schema-less documents.
Node.js Applications
Embedded database for Node.js apps requiring local storage. No external dependencies, no server setup, no compilation. Works on all platforms without native bindings.
Desktop & CLI Tools
Perfect for desktop apps (Electron, Tauri) and CLI tools. Store configuration, cache data, manage local state—all with npm install.
Rapid Prototyping
Skip database setup entirely. Query with JavaScript objects, not SQL strings. Handles 10K-500K documents with intelligent caching. Migrate to PostgreSQL or MongoDB when you scale.
Embedded Systems
Local-first applications, IoT devices, edge computing. Single-instance architecture with file-based storage. Built-in GUI for data inspection during development.
AxioDB is not competing with PostgreSQL or MongoDB. It's for when you need a database embedded in your app—Electron, CLI tools, local-first apps. Sweet spot: 10K-500K documents. No native dependencies, no server setup.
The Problem With the Usual Options
SQLite:
- ✗ Requires native C bindings (better-sqlite3, node-sqlite3)
- ✗
electron-rebuildon every Electron update, platform-specific compilation - ✗ SQL strings instead of JavaScript objects
- ✗ Schema migrations when your data model changes
JSON Files:
- ✗ Full file read/write for every operation
- ✗ No built-in querying, indexing, or caching
- ✗ Linear O(n) search performance
MongoDB (Server):
- ✗ Requires a separate server process
- ✗ Overkill for small-to-medium, single-app datasets
- ✗ Not suitable for embedded/desktop scenarios
AxioDB is pure JavaScript, embedded, with MongoDB-style queries built in:
- ✓Works everywhere Node.js runs—no rebuild, no native dependencies
- ✓MongoDB-style queries:
{age: {$gt: 25}} - ✓Schema-less JSON documents—no migrations
See the full feature-by-feature comparison against SQLite, JSON files, lowdb, nedb, and better-sqlite3.
