Lab of PDF

Free online PDF tools — merge, split, compress, watermark, and convert PDFs entirely in your browser. No document uploads or sign-up; practical capacity depends on the device and browser.

Use the production tools at labofpdf.com, make a large scan smaller with the private scanned PDF compressor, read why scanned PDFs become large and how to reduce them, or start with the printable PDF submission checklist before uploading a document to a university, visa, government, job, grant, or client portal.

Features

Tool	Description
Merge PDF	Combine multiple PDFs into one. Preview pages and pick exactly which ones to include.
Split PDF	Extract specific pages from a PDF or split it into two separate files.
Compress PDF	Reduce PDF file size losslessly — text stays selectable.
Compress Scanned PDF	Detect image-based scans, choose a size target, and make the document smaller locally.
PDF to Word	Convert PDF to Word document. OCR applied automatically to scanned pages.
Add Watermark	Add a text watermark to every page of your PDF.
Remove Watermark Annotations	Inspect supported Stamp/Watermark annotations and remove only selected candidates.
Manage Pages	Reorder, rotate, remove, restore, duplicate, extract, or insert blank pages with a visual preview.
PDF to Image	Convert PDF pages to PNG or JPEG images. Download individually or as ZIP.
Images to PDF	Arrange JPG, PNG, or WebP images visually and create A4, Letter, or image-sized PDF pages.
Visual PDF Editor	Drag, resize, align, layer, lock, hide, copy, sign, highlight, draw, and securely redact objects on real PDF pages.

All processing happens locally in your browser using pdf-lib and pdfjs-dist. Your files never leave your device.

Tech Stack

• React 19 + TypeScript
• Vite (build tool)
• TailwindCSS 4 (styling)
• pdf-lib / pdfjs-dist (PDF processing)
• Tesseract.js (OCR)
• JSZip (ZIP downloads)
• Zustand (state management)
• lottie-web (animations)
• Lucide (icons)

Getting Started

npm install
npm run dev

Open http://localhost:5173 in your browser.

Build

npm run build

Output goes to dist/.

The production build also creates route-specific HTML files such as dist/to-image.html. Configure Nginx to resolve $uri.html before the SPA fallback so search engines and link-preview bots receive the correct title, description, canonical URL, and structured data without running JavaScript.

Nginx production configuration

Use deploy/nginx-labofpdf.conf as the application portion of the HTTPS server block. It includes:

• route-specific SEO HTML resolution;
• one-year immutable caching for hashed assets;
• no-cache handling for the application entry document;
• gzip compression;
• HSTS, permissions policy, MIME sniffing and framing protection;
• a report-only Content Security Policy that can be monitored before enforcement.

Run nginx -t before reloading Nginx. Keep the CSP in report-only mode until PDF conversion, OCR, analytics, social sharing, and downloads have been tested on the deployed domain.

Server disk usage

Deploy into one fixed working directory instead of creating a permanent copy for every release. Vite clears dist/ before rebuilding, so old hashed assets do not accumulate. Once the build has succeeded and Nginx is serving dist/, node_modules/ can be removed; restore it with npm ci for the next build. If release directories are used for rollback, retain the active release and at most two previous releases.

See PROJECT_HANDOFF.md for the current deployment state and the complete cross-agent verification checklist.

Canonical deployment chain

GitHub main → canonical local worktree → lint and tests → npm run build → validated dist/ → atomic upload → Nginx /var/www/labofpdf.

Run deploy/deploy-to-server.sh production-YYYY-MM-DD-description only from the canonical worktree after pushing that new immutable tag. The build writes dist/release.json, and deployment succeeds only when the online manifest reports the same Git commit. The deployment also installs the local Node API at /opt/labofpdf-api and restarts visitor-counter.service on port 3001 behind /api/. API state is kept in /var/lib/labofpdf, outside the static release directories.

Privacy

All PDF processing runs entirely client-side. No files are uploaded to the server. After a result is downloaded, users can optionally report whether it worked. Feedback stores the tool path, outcome, reason, optional short comment, and release commit in a server-local SQLite database. It does not store the PDF, file name, document content, or IP address.

Administrators can view privacy-safe 30-day aggregates on the server with:

sudo -u www-data node /opt/labofpdf-api/feedback-report.mjs

License

MIT