PDF Tools
PDF guide

How to read the text in a scanned PDF (OCR)

Keep the original pages and add an invisible text layer for search and selection, or download the recognized wording as text. Browser mode can read scanned areas on pages that already contain selectable text.

Updated . Tested step by step on this tool.

How to use PDF OCR

  1. Choose your scanned PDF, then select its language.
  2. Choose Searchable PDF to preserve the original pages with an invisible text layer, or Text file to save only the recognized wording.
  3. Choose Standard to run locally, or Advanced with a free account to run on the server. Browser OCR downloads the selected language data on first use and caches it on this device.
  4. In Standard mode, keep Keep existing selectable text and read scanned areas checked for mixed pages. For Text file output, optionally enable Straighten scanned text for recognition when the scanned wording is tilted.
  5. For Searchable PDF output in Advanced mode, Skip pages with selectable text in searchable PDF leaves those pages without a new OCR layer. Turn it off if a native label shares the page with a scanned body that still needs a searchable layer. Advanced Text file output recognizes every page.
  6. Select Read the text in this PDF. Review the wording and the browser mode Recognition confidence list, then download your result. A searchable PDF can continue into another PDF tool.
Browser OCR keeps native text, reads scanned areas, and reports recognition confidence for each page. Searchable PDF keeps the original page appearance.
Browser OCR keeps native text, reads scanned areas, and reports recognition confidence for each page. Searchable PDF keeps the original page appearance.

Before you export

Recognition can contain errors. Searchable output preserves page size, rotation and original pixels; it adds selection and search rather than changing the visible scan.

Recognition confidence is the engine's estimate, not a guarantee of accuracy. Browser mode flags pages below 70% for review and reports pages where it kept existing selectable text. Check names, numbers and important wording against the original, including pages with higher confidence.

Browser mode can be stopped with Cancel OCR. Advanced mode uses Stop waiting, which leaves the server job running. A canceled browser run can be started again with the same source file.

The workflow controls can guide you from a scan to searchable output and review, or from a redacted copy to searchable output and password protection. Export at each stage, then use Continue to the suggested next tool. Saving a workflow remembers its sequence on this device; it does not save the PDF or passwords.

Frequently asked questions

What happens to pages that already contain text?

In Standard browser mode, Keep existing selectable text and read scanned areas preserves native text while recognizing remaining image areas. A native heading does not cause the scanned body to be skipped. Pages that contain only native text need no OCR. If existing text cannot be located safely, the page is kept and a note explains the limit. For Searchable PDF output in Advanced mode, the skip-pages option leaves pages with existing selectable text without a new OCR layer. Advanced Text file output recognizes every page.

Does straightening change my PDF pages?

Straighten scanned text for recognition is available in browser mode for Text file output only. It adjusts a detected tilt up to 15 degrees during recognition. Searchable PDF keeps the original page orientation and text positions, and does not use this straightening option.

Can I stop OCR?

Cancel OCR stops browser recognition and keeps the source file available. In Advanced server mode, Stop waiting stops waiting in this page; it does not cancel the server job, and credits may still apply.

Do browser jobs upload my PDF?

Standard processing keeps the PDF on your device. Advanced server processing uploads it for that job and requires a free account.