Searchable PDF
Recognize a scanned technical record and add a selectable English, Spanish or Russian text layer, within a 10-page, 10 MiB limit per file. Files stay in your browser.
Recognition files ≈ 26 MB on first use, plus language data.
Recognized text
Was this tool useful?
Try an example
Download sampleSelect English. Download the PDF and search for INVOICE, 148.50, DELIVERY and 2048. Both pages should look like the sample, with text you can select and copy.
What to expect
Each page is recognized as an image, in order. English, Russian and mixed recognition are available. OCR can misread letters, numbers, columns and tables: review the result. This creates a new PDF from rendered page images and an invisible text layer. Visible page size and rotation are retained; forms, links, bookmarks, attachments and digital signatures are not retained. Existing text is recognized again; image quality is limited by rendering resolution.
Pages are rendered up to 4 megapixels each. Loading and each processing step allow up to 90 seconds; the whole job stops after 10 minutes. Blank pages are marked in the text. No download is offered if processing fails or no text is recognized.
Protected PDF?Unlock it with its password first.
The first run may download converter files and take a moment. If it fails, check your connection or try a smaller file. Cancel stops active processing; you can choose another file and retry.
Engine licenses and source code
A practical guide
Add a text layer to the archived scan.
- Choose the scanned record PDF, within 10 pages and 10 MiB.
- Select English, Spanish or Russian for recognition.
- Search the result for a known identifier to verify the text layer.
Try this example
Recognize a 6-page scanned inspection log in Russian; searching for an equipment ID on page 5 should locate it in the resulting PDF.
Before you start: Recognition is limited to English, Spanish and Russian, 10 pages and 10 MiB; small print and low scan quality can reduce accuracy.