Lumina

AI-Powered Document Search

Early Access Beta Built with Transformers.js Linux Only

Welcome to your document search companion

Import your documents and use AI to find exactly what you need, with your data staying private on your device.

Lumina UI Screenshot

Import

Search

Discover

Features

  • Lightning Fast Search - Search your entire document database in less than 4 seconds
  • 100% Local & Private - All processing happens on your device
  • Supported File Formats - Currently supports MD, DOCX, PDF, ODT, TXT, and HTML
  • Semantic Understanding - Find documents based on meaning, not just keywords

How It Works

  1. Import Your Documents - Point Lumina to your document folders or drag and drop files
  2. Initial Embedding - Lumina processes and embeds your documents (takes just a few minutes)
  3. Search Naturally - Ask questions in natural language and get relevant document results instantly

What's Next?

  • Custom AI model finetuning
  • Plugin system for Markdown/PDFs

Installation

Current version: 1.0.0 | Released: April 9, 2025

Requirements

  • Linux distribution with FUSE support
  • 4GB RAM minimum (8GB recommended)
  • 300MB free disk space

Instructions

# Make the AppImage executable
chmod +x ./Lumina-1.0.0.AppImage

# Run Lumina
./Lumina-1.0.0.AppImage

FAQ

Is my data sent to the cloud?

No. Lumina processes all your documents locally on your machine. Your data never leaves your computer.

How long does the initial embedding take?

The initial embedding process typically takes a couple of minutes depending on the number and size of your documents.

What document formats are supported?

Lumina currently supports the following formats: Markdown (MD), Word Documents (DOCX), PDF, OpenDocument Text (ODT), plain text (TXT), and HTML.

Do I need an internet connection?

No. Once installed, Lumina works completely offline.

Is Lumina open source?

Lumina is a closed-source project. No ads, no telemetry, no BS.