FormatDogWorkspace

100% SECURE PRIVATE SANDBOX

Transparency

"Everything runs in your browser, nothing is uploaded" is an easy claim for any site to make. This page exists so you don't have to take it on faith — it names the exact open-source software doing the work, and gives you a way to prove the no-upload claim yourself in about ten seconds.

Prove it: the Airplane Mode test

Turn off your internet connection, then use any tool on this site. Merge a PDF, convert an image, run OCR — whatever you like. If FormatDog were secretly uploading your files, every one of those would fail the instant your connection drops. They don't, because nothing is ever sent anywhere.

This live badge reads your browser's actual navigator.onLine connection status — toggle airplane mode or unplug your network and watch it flip to offline in real time, right here on this page.

What's actually running under the hood

FormatDog isn't a thin wrapper around a mystery server-side "engine." Every conversion, compression, OCR pass, and image edit is done by a named, established open-source library, compiled to run directly in your browser via JavaScript or WebAssembly. Here's the real list:

PDF.js

Reads and renders PDF pages, and extracts real text content, entirely client-side.

Mozilla · Apache License 2.0 · github.com/mozilla/pdf.js

pdf-lib

Creates, merges, splits, and edits PDF documents — the core engine behind most PDF tools here.

MIT License · github.com/Hopding/pdf-lib

docx

Builds real, valid .docx Word files — used by PDF to Word and related converters.

MIT License · github.com/dolanmiu/docx

Mammoth.js

Reads .docx Word files for Word to PDF conversion.

BSD 2-Clause License · github.com/mwilliamson/mammoth.js

SheetJS (xlsx)

Reads and writes Excel spreadsheets for the Excel converters.

Apache License 2.0 · github.com/SheetJS/sheetjs

PptxGenJS

Builds PowerPoint files for the PDF to PowerPoint converter.

MIT License · github.com/gitbrent/PptxGenJS

Tesseract.js

Runs real OCR (optical character recognition) on scanned pages and images, fully offline.

Apache License 2.0 · github.com/naptha/tesseract.js

qpdf

Compiled to WebAssembly to handle PDF encryption, decryption, and structural repair.

Apache License 2.0 · github.com/qpdf/qpdf

OpenCV.js

Handles image-processing tasks like document edge detection in the Doc Scanner tool.

Apache License 2.0 · github.com/opencv/opencv

TensorFlow.js & body-pix / body-segmentation

Detects a person in a photo for the background-removal tool — the model runs locally; your photo is never sent anywhere to be analyzed.

Google · Apache License 2.0 · github.com/tensorflow/tfjs

heic2any

Converts iPhone HEIC/HEIF photos to standard formats.

MIT License · github.com/alexcorvi/heic2any

JSZip

Reads and writes ZIP-based file formats (PowerPoint files are ZIP archives internally).

MIT License · github.com/Stuk/jszip

qrcode.js

Generates the QR code matrix behind the QR Code Studio.

MIT License · by Kazuhiko Arase

Every one of these is a well-known, independently auditable open-source project — none of them phone home, and none of them are proprietary code we're asking you to trust blindly. If you'd like to verify any claim on this page yourself, your browser's developer tools (Network tab) will show you exactly what does, and doesn't, leave your device while you use any tool here.

Something not add up?

If you spot a library we've missed crediting, or a claim on this page that doesn't match what you observe, please let us know — this page is meant to be accurate, not just reassuring.