Troubleshooting the Online Editor

Fixes for common problems in the online editor: page reloads on first open, slow model downloads, nothing found on scans, sign-in prompts, limits and download issues.

The editor reloads once when I open it

This is expected. On first load the editor may reload itself once to turn on the browser isolation it needs to run AI models and OCR on several CPU threads. If your browser or an extension blocks that, the editor still works, but more slowly.

Detection is slow, or stuck at the start

The first time you use Local NER, OCR or face detection, the browser downloads the model, which can be several hundred megabytes. Later runs use the cached copy.
Local NER runs on your CPU. Try a smaller model, raise CPU threads in OCR settings, or switch back to Cloud PII.
Click the progress ring in the top bar to see the per-page log. Pages that failed have a retry button.

Nothing was found on a scanned page

Check the page's status line. OCR should show a green tick.
Make sure the OCR language matches the document's script, see OCR settings.
If the PDF has a text layer that is garbled, turn on Always run OCR and run detection again.
Blurry, rotated or handwritten text may not be read. Draw boxes over it by hand.

Some names or numbers were not detected

Under PII Filter, check that the type is not switched off (an outlined chip).
With Local NER, lower the Detection threshold, or try another model.
Add Rule-based detection for structured numbers such as IBANs, card numbers and SSNs.
Select the missed text and choose All to redact every occurrence, or use Search in the Manual Redaction tab.

Too many false positives

Click the type chip to exclude a whole type, or remove single items from its list. With Local NER, raise the Detection threshold. With rules, raise the Minimum score.

Downloading the redacted file needs a free account. Sign in with Google, Microsoft or an e-mail link, then click Save PDF again. Only the built-in sample document can be saved without signing in.

My file will not open

The editor accepts PDF files only.
Check the file size: 5 MB without an account, 25 MB on Free, 50 MB on paid plans.
For a password-protected PDF, enter the document password when asked.
Damaged PDFs may fail to load. Try printing the file to a new PDF first.

Only part of my document was processed

Pages beyond your per-document limit are not scanned: 10 without an account, 25 on Free, 100 on paid plans. Split the document or upgrade. See Usage and limits.

"Page limit reached for this billing period"

You have used your plan's pages for this month. Upgrade, top up credits or wait for the next period. See Plans and billing.

The Google Drive or Dropbox picker does not open

The links are in the Start Redacting Your PDF dialog. The picker opens in a sign-in window, so allow pop-ups for pdf-redaction.com if your browser blocks it. See Google Drive integration.

The saved PDF looks different from the original

Saved pages are flattened images at A4 size with a searchable text layer, so form fields are no longer editable and the file can be larger. This is how the editor makes sure covered text is really removed. See How it works.

Still stuck?

Open the chat with the support button in the bottom bar of the editor, or create a support ticket. Include the file name, the page number and what you expected to happen. Do not attach documents with real personal data.

Troubleshooting