How PDFExtract Works Fully On-Device on iPhone

TL;DR

Every step in PDFExtract — importing PDFs, extracting images and text, and OCR — runs locally on your device, so your documents are never sent to a server.

Why it’s fully on-device

PDF import happens locally — files aren’t uploaded.
Image extraction runs on the device.
Text extraction and Markdown generation are local.
OCR fallback for scanned pages also runs on-device.

Why it works this way

The documents you extract from — contracts, statements, research, scanned records — are often confidential, and routing them through a conversion server is a real exposure. PDFExtract keeps the whole pipeline on your device, including OCR, so there’s no upload to intercept and nothing retained elsewhere. Local isn’t a mode here; it’s the design.

Tips & edge cases

Airplane mode works — extraction and OCR don’t need a connection.
Confidential PDFs stay on your phone end to end.
No account means nothing tied to your documents.

FAQ

Is any part of extraction done in the cloud? No. PDF import, image and text extraction, and OCR fallback all run locally on your device.

Are my documents uploaded? No. Your files never go to a server.

Does that include scanned documents and OCR? Yes. Even OCR of scanned pages happens on-device.