🌙← Back to home

Privacy

Last updated: June 19, 2026

PDF Dark is built around one core promise: your PDF file never leaves your browser. The file you drop in — its name, content, text, and images — stays on your device. The site itself (this page, page-view analytics, and the ads we may show to keep the tool free) works like any normal free utility — details below.

Your PDF is processed locally

When you drop a PDF onto this page, the entire conversion happens inside your browser:

  1. Your browser reads the file locally via the File API.
  2. PDF.js parses and renders each page to a canvas.
  3. Our dark-mode algorithm runs in a Web Worker — still inside your browser.
  4. On Download, your browser assembles the themed PDF and saves it locally.

At no point is your PDF uploaded to a server. You can verify this by opening the browser's Network tab during use — you'll see no file upload request. The file name, content, text, and images inside the PDF all stay on your device.

What we do not do

  • We don't upload, store, or read your PDF file — name, content, and extracted text all stay on your device.
  • We don't ask for your email, log you in, or build a long-term profile tied to you.
  • We don't correlate the file you converted with anything else — not with the analytics events, not with the ads, not with any third party. Your file is processed locally and forgotten.
  • We don't sell your data. We don't have a database of user data to sell.

Site analytics

To understand which pages people find useful and where the site is slow or broken, we run two analytics tools — both on the site itself, never on your PDF file:

  • Google Analytics 4 — aggregate page views, referrers, country, device type. IP addresses are anonymized by GA4 by default and not stored.
  • Microsoft Clarity — anonymous heatmaps and session recordings of how people interact with the page (clicks, scrolls). Clarity masks text and inputs by default, so the content of your PDF and any text you type are not captured.

Neither tool sees your PDF file. If you'd rather not be counted, a browser-level tracker blocker (uBlock Origin, Brave Shields, etc.) will block both.

Ads

PDF Dark is free and runs on a paid hosting bill, so we plan to show display ads on the site to cover costs. We'll likely use Google AdSense (the standard publisher network), which means:

  • Ads appear on the site, in clearly marked slots — never inside the PDF you download, and never on top of the converter in a way that gets in the way of the tool.
  • Like all AdSense placements, ads may use cookies for frequency capping, fraud prevention, and basic targeting. You can opt out of personalized ads at Google's Ad Settings, or block them entirely with a tracker blocker.
  • Ad networks do not receive your PDF file or anything extracted from it — your file never leaves your browser, so there's nothing to send.

When ads go live, we'll bump the “Last updated” date and list the exact networks here so you know what cookies to expect.

Error monitoring

To catch bugs we'd otherwise never hear about, we run Sentry for error monitoring. Sentry only activates when something actually breaks (e.g. a PDF fails to parse, the Web Worker crashes). It then sends:

  • The JavaScript stack trace of the error
  • Browser and operating system version
  • A short, masked session replay of the failure — text, inputs, and the rendered PDF are all hidden so the replay shows structure but not content

No cookies are set. No data is sent for users who don't encounter an error. IP addresses are not stored. Data is processed by Sentry per their privacy policy.

Hosting

This site runs on Vercel. Like any web host, Vercel keeps short-term HTTP access logs (IP address, URL, user agent) needed to serve the site and absorb attacks. These logs are not used for analytics or profiling and are handled per Vercel's privacy policy.

Changes

If anything material changes — a new analytics or ad network, a new third-party script, anything that touches the “your file never leaves your browser” promise — we'll update this page first, bump the “Last updated” date above, and disclose what changed before the new code ships.

Questions or data requests

Since we don't have accounts or profiles, there's almost nothing tied to you personally — the traces we have are aggregate Google Analytics and Clarity events, ad-network cookies (once ads are live), short-term Vercel access logs, and (if something crashes) a masked Sentry error report. If you want to check, correct, or delete anything under GDPR / UK / CCPA, or just have a question, email hello@pdfdark.org.