kipeles/apexkit-releases
ApexKit The AI-Native, Multi-Tenant Backend in a Single Binary. Built with Rust, SQLite, and React. Official Website & Docs • Features • Downloads • Quickstart • Architecture What is ApexKit? ApexKit is a blazingly fast, all-in-one backend solution designed for the AI era. Inspired by tools like PocketBase and Supabase, ApexKit compiles your entire backend infrastructure—Database, Authentication, Storage, Real-time APIs, Serverless JS Scripting… See the full description on the dataset page: https://huggingface.co/datasets/kipeles/apexkit-releases.
<p align="center"> <img src="static/images/apexkit-logo.svg" width="250" alt="ApexKit Logo"> </p>
<h1 align="center">ApexKit</h1>
<p align="center"> <strong>The AI-Native, Multi-Tenant Backend in a Single Binary.</strong><br> <em>Built with Rust, SQLite, and React.</em> </p>
<p align="center"> <a href="https://apexkit.pages.dev"><strong>Official Website & Docs</strong></a> • <a href="#features">Features</a> • <a href="#downloads">Downloads</a> • <a href="#quickstart">Quickstart</a> • <a href="#architecture">Architecture</a> </p>
What is ApexKit?
ApexKit is a blazingly fast, all-in-one backend solution designed for the AI era. Inspired by tools like PocketBase and Supabase, ApexKit compiles your entire backend infrastructure—Database, Authentication, Storage, Real-time APIs, Serverless JS Scripting, and a fully featured Admin Dashboard—into a single, portable Rust binary.
Beyond standard BaaS features, ApexKit introduces natively integrated Generative AI (Gemini/Gemma), Vector Semantic Search, and an AI Architect that can literally build your schemas and server scripts via natural language chat.
✨ Key Features
🗄️ Database & Auto-APIs
- Embedded SQLite: High-performance data storage with WAL mode, batching, and zero-copy JSONB processing.
- Auto-generated APIs: Instant REST and dynamic GraphQL endpoints for every collection you create.
- Advanced Query Engine: Deep filtering, sorting, joins (relations), and pipeline aggregations natively supported.
🧠 AI-Native Capabilities
- AI Architect: Chat with the built-in AI assistant from the dashboard to auto-generate collections, write server-side JavaScript, and design UI templates.
- Vector Search Engine: Natively embedded local vector database using
candleand HNSW. Automatically embed text and images for semantic search. - Instant Search: Typo-tolerant, ultra-fast full-text search powered by Tantivy.
- AI Actions: Securely define and execute LLM prompts on the server, exposing them as simple API endpoints to your frontend.
⚡ High-Performance Edge Runtime (QuickJS + TS/WASM)
- Embedded QuickJS Engine: Lightning-fast, standards-compliant JavaScript/TypeScript runtime powered by QuickJS (
rquickjs) with full async/await and ES Module support. - Native TypeScript & TSX: First-class support for
.tsand.tsxscripts (via in-memory Oxc compiler). Run webhooks using Hono with zero compile steps. - WASM & WASI Orchestration: Call high-performance WebAssembly binaries and precompiled modules directly from scripts using
$wasm. - Preemptible Task Scheduling: Built-in quantum round-robin scheduler ensures long-running loops or tasks yield CPU slices without freezing Tokio worker threads.
- Rich Standard Globals: Built-in access to
$db,$http,$files,$fs,$cache,$mail,$realtime,$ai, and$env.
🏢 Built-in Multi-Tenancy & Sandboxes
- Designed for B2B/SaaS apps out of the box. Securely isolate data using Tenants.
- Instantly spin up Sandboxes to safely test schema changes or let the AI Architect draft features without touching production data.
📡 Real-time & Storage
- WebSockets & SSE: Subscribe to database changes or broadcast custom ephemeral events to channels.
- S3 & Local Storage: Upload files, generate thumbnails on the fly (AVIF/WebP), and seamlessly migrate data between local disks and AWS S3/Cloudflare Spaces.
- Static Hosting: Deploy your frontend (React, Vue, HTML/CSS) directly into ApexKit for a true all-in-one deployment.
🚀 Quickstart
1. Run via Docker (Recommended)
The easiest way to get started is using our lightweight Docker image (Alpine/Musl based).
docker run -p 5000:5000 \
-v /path/to/local/data:/app/storage \
deniskipeles/apexkit:latest2. Run via Pre-compiled Binary
<a id="downloads"></a> Download pre-built standalone binaries for your target architecture from our global CDN:
Verify download integrity using the official [SHA-256 Checksums](https://huggingface.co/datasets/kipeles/apexkit-releases/resolve/main/v0.1.0-beta.1/checksums.txt).
# Linux / macOS
tar -xzf apexkit-*.tar.gz
chmod +x apexkit
./apexkit --port 5000
# Windows
# Extract the zip archive and run:
apexkit.exe --port 50003. Access the Admin UI
Once the server is running, navigate to: 👉 `http://localhost:5000/_dashboard`
Log in using the default credentials:
- Email:
admin@apexkit.io - Password:
password(Change this immediately in the Users tab!)
🏗️ Architecture & Tech Stack
ApexKit achieves its performance and small footprint by utilizing a modern, carefully selected tech stack:
- Official Website: apexkit.pages.dev
- Core Backend: Rust & Axum
- Database Engine: SQLite (via
rusqlitewith zero-copy JSONB and WAL mode) - Search & Vector Engine: Tantivy (Full-text search), Candle & ONNX Runtime
- JavaScript & TS Runtime: QuickJS (via
rquickjs) & Oxc (High-speed TypeScript transpiler) - WebAssembly Engine: Wasmtime (WASI Preview 1)
- Admin Dashboard: React 19, Vite, Tailwind CSS, & Monaco Editor
📚 Documentation
Detailed guides for interacting with the SDK, writing server-side JavaScript, and configuring AI models can be found inside the Admin UI by clicking the "API Docs" button on the Collections or Users pages, or by visiting the /scalar endpoint on your running instance for OpenAPI specifications.
Basic SDK Usage Example (JS/TS)
import { ApexKit } from '@apexkit/sdk';
const pb = new ApexKit('http://localhost:5000');
// Authenticate
await pb.auth.login('user@example.com', 'password123');
// Create a record
const post = await pb.collection('posts').create({
title: "Hello World",
content: "Building with ApexKit is awesome!"
});
// Semantic Vector Search
const similarPosts = await pb.collection('posts').searchTextVector("Greeting messages", 5);🛠️ Building from Source
To build ApexKit from source, you will need Rust, Node.js (for the Admin UI), and a few system dependencies (like cmake and nasm).
# 1. Clone the repository
git clone https://github.com/deniskipeles/apexkit.git
cd apexkit
# 2. Build the Admin UI
cd admin-ui
npm install
npm run build
cd ..
# 3. Move UI assets to the static directory
mkdir -p static/dashboard
cp -r admin-ui/dist/* static/dashboard/
# 4. Build the Rust backend (Release mode)
cargo build --release -p apexkit-api
# 5. Run it!
./target/release/apexkit-api⚖️ License & The "Anti-Big-Tech" Guarantee
This engine is licensed under the Business Source License (BSL 1.1). See the full LICENSE file for legal details.
The TL;DR for Startups & Devs:
- ✅ Building an App/SaaS? 100% Free forever. Self-host it on your own servers, scale globally, and keep 100% of your revenue.
- ✅ Building a Competing Service? If you want to use our engine to launch your own hosting platform or managed BaaS to compete with us, go for it! As long as your company makes under $5M/year or has raised under $10M, you have our full blessing. Work hard, innovate, and keep us on our toes.
- ❌ Are you a Tech Giant? If you are a mega-corporation (like AWS, GCP, Azure) or a company making over $5M/year, you cannot bundle or host our engine as a managed service without a separate commercial agreement.
No Vendor Lock-In
Every line of code we write automatically converts into pure, unrestricted Apache 2.0 open-source exactly 2 years after its release. If we disappear or stop maintaining the project, the community owns the code under Apache 2.0 forever.
