Privacy Architecture
DearPDF is built so contracts, invoices, and scans do not need to leave this device. Processing is intentional local compute — not a privacy mode bolted onto an upload queue.
Typical cloud converters
- Multipart upload of your raw document to remote servers
- Temporary files written on someone else’s disk
- Download links and logs that can linger
Your document leaves your device and may be retained longer than you expect.
DearPDF client-side engine
- Direct file open in this browser tab
- Local processing for merge, compress, OCR, and more
- Instant local download — no remote staging URL required
Built so documents do not need to be uploaded for conversion.
Four layers of isolation
Defense-in-depth guarantees for every DearPDF tool.
Browser sandbox
PDF tools run inside your browser process. Document data is not handed to a remote conversion queue.
Ephemeral working copy
Files are held in tab memory while you work. Closing the tab clears that working copy.
Offline-capable after first visit
Cached app assets let core tools keep working without an active connection.
No document telemetry
Filenames and file contents are never sent. Optional anonymous usage pings (tool name, event type) may be used to improve the product.
- PDF and image processing happen in your browser tab.
- Files are not uploaded to DearPDF servers for conversion.
- Passwords for encryption stay in memory on this device.
- OCR language models may be downloaded from public CDNs by Tesseract.js into your browser cache — your document itself is not sent to DearPDF.
- No account is required to use the tools.
- Anonymous usage pings may be sent (tool name, event type, time, random session id) so we can see which tools are used. Filenames and file contents are never sent.
Always verify the downloaded result before sharing sensitive documents. How it Works.