What the browser does with your file

The short version is easy to over-read. The file you select is processed in the tab and is not stored by this site. That does not mean the page works with the network unplugged, and it does not mean every byte of the page came from your computer.

Two different downloads

The tool page loads JavaScript libraries from a public CDN. PDF tools use pdf-lib and, where a page has to be painted or its text read, pdf.js. HEIC to JPG loads heic2any from jsDelivr. English OCR loads Tesseract when you turn that option on. Those scripts are not your file. If they fail to load, the buttons cannot run. Refresh on a working connection.

The file you pick is a separate step. The browser reads it with the file picker, in memory, in that tab. Compress, merge, split, Word conversion, and the image tools use those bytes locally. This site does not receive a copy to store, and there is no account attached to the file.

Closing the tab drops the in-memory copy. Download anything you want to keep. A later visit does not have the previous file.

What still needs the network

  • The HTML, the CSS, and the CDN libraries, the first time you open the tool.
  • English OCR, which fetches its engine when you enable it.
  • HEIC conversion, which needs the heic2any script.
  • The page will not run OCR if you open a saved HTML file as file://. Digital text extraction can still work in that situation. OCR needs the live site.

What this does not mean

It does not mean a password can be stripped. Protected PDFs are refused on the PDF tools. It does not mean the result equals Adobe, Word, or a desktop HEIC converter. Each tool page states the specific limit: 40 pages for compress, 80 for the text conversions, 60 for organize, fixed JPEG quality for HEIC, principal and interest only for the mortgage calculator.

It does not mean the same sentence needs to be the content of every tool. The limits above are the useful part. The full account of what the operator does with personal data is on the Privacy page, not in a repeated line under every button.

If a result looks wrong, the usual cause is the limit, not a hidden upload. A text PDF that grew was rasterized. A Word file full of question marks was drawn in Helvetica. An empty Excel sheet was a scan. The guides linked from the home page go through those cases one by one.

The mortgage calculator does not read a file at all. The numbers you type stay in the page. The same rule about not storing a visitor's document does not turn that calculator into financial advice. It is still principal and interest only, labeled in dollars.

Privacy, the operator, and what the site itself logs are written on the Privacy page and the Terms page. Those pages are the place for that account. A tool guide is the place for the limit of the tool: page caps, missing fonts, lost transparency, and text that will not survive compression.

Questions

Is the PDF uploaded to PC Tools Online?

No. The file is read in the tab and is not stored here. The libraries that run the tool are loaded from a CDN.

Will it work offline?

Not on a first visit. After the scripts have loaded, a digital PDF can be handled in the tab. OCR and HEIC still depend on their scripts being available.

Why does a double-clicked HTML file refuse OCR?

Chrome blocks that worker on file://. Open the tool on the live site.