The short version: PDFMend does not upload your PDF to a PDFMend server. The document, map, and variants are processed and stored in your browser. Text you choose to translate may be sent to the translation provider you select.
1. What PDFMend is
PDFMend is a browser-based PDF editor for mapping pages, translating text, reconstructing content, and creating selectable text layers. It is a static web application. There is no PDFMend account system and no PDFMend backend that receives your document.
2. Your PDF and project data
When you open a PDF, the file is read locally by your browser. Region maps, text items, images, and content variants are kept in browser storage so the session can survive a refresh. They are not sent to us.
You can remove the browser copy at any time with File ▸ Forget saved session or by clearing site data in your browser settings. Exported files are written by your browser to the location you choose.
3. Browser storage
PDFMend uses browser storage, including IndexedDB and localStorage, for:
- the open PDF and its session state;
- region maps and content variants;
- preferences such as language, OCR, and translation settings.
This storage is origin-specific. It remains on your device until you clear it, use the forget-session command, or your browser removes it. PDFMend does not use advertising cookies or sell data.
4. Translation providers
Translation is optional. When you translate text, the selected text is sent to the provider you choose:
- Chrome on-device translation runs locally in a supported Chrome browser; text is not sent to a remote translation service.
- LibreTranslate or a custom endpoint receives the text you submit, and possibly supporting language parameters, in order to return a translation. That provider's own privacy policy and retention practices apply.
- Manual / external workflows leave translation to tools you operate yourself.
Do not translate confidential text through a third-party server unless you are comfortable with that provider's handling of it. PDFMend does not control those servers.
5. OCR and detection
OCR is performed in your browser with Tesseract.js. OCR language data and the engine may be downloaded from a public CDN on first use and cached by your browser. Text recognition and detection results stay local; PDFMend does not receive the page images or recognized text.
6. Third-party requests
The app loads open-source libraries from public CDNs and may contact the translation endpoint you select. Those requests necessarily expose standard technical information such as your IP address, browser type, and the requested URL to the relevant provider. PDFMend does not include analytics, advertising, or behavioral tracking.
7. Data retention
We do not retain a copy of your document or project because we do not receive one. Local browser storage remains under your control. Third-party translation providers may retain submitted text according to their own policies.
8. Security
PDFMend is served over HTTPS in production. Local-first processing reduces the amount of document data that leaves your device, but no browser application can guarantee absolute security. Keep your browser and operating system updated, and avoid opening untrusted PDFs.
9. Children
PDFMend is a general-purpose document tool and is not directed at children under 13. Do not use it to process a child's personal information without appropriate consent.
10. Changes
We may update this policy as the app changes. The effective date at the top of this page will change when we do. Material changes will be reflected in the public repository and, where appropriate, in the app.
11. Contact
For privacy questions, open a thread in GitHub Discussions or use the contact details in the repository profile. Please do not post confidential documents in a public issue or discussion.