Make a scanned PDF searchable
OCR PDF on ilovspdfs recognises the text in a scanned PDF and writes it as an invisible layer over each page, so the document can be searched and its text copied while looking exactly as before. English only for now, up to 20 pages a run. Part of Pro: sent over HTTPS, processed in memory, never stored.
Pro tool. The file is sent over HTTPS, processed in memory, and never stored.
Drop your file here
.pdf · up to 4.0 MB
No files chosen yet
About this tool
A scan is a photograph of a page: your reader cannot search it, and selecting text does nothing. OCR looks at the picture, works out the letters, and tucks an invisible text layer behind the image so the document reads normally and searches properly. Nothing visible changes. Pages that already carry text are left alone and counted in the result. Recognition is in English for now, and accuracy follows the quality of the scan: a crisp 300 dpi page comes out nearly perfect, a skewed phone photo does not.
How to make a scanned PDF searchable
- 1Choose the scanned PDF.
- 2Run recognition.
- 3Download the searchable PDF and try a search in your reader.
At a glance
| What it does | Make a scanned PDF searchable with an invisible text layer. |
|---|---|
| Where it runs | Sent over HTTPS, processed in memory, never stored |
| Accepts | |
| Files per job | One at a time |
| Maximum file size | 4 MB |
| Maximum pages | 20 |
| Price | Part of Pro. Needs an account and a subscription; no watermark. |
| Status | Working |
Questions
- Is my file uploaded for this tool?
- Yes. Unlike the tools marked Private, this one sends your file to an ilovspdfs server over HTTPS, processes it in memory, and streams the result straight back as the response. Nothing is written to disk, nothing is logged about the contents, and nothing remains once the response has been sent.
- Which languages are supported?
- English only at the moment. Other languages need their own recognition data, and we would rather add them one at a time than claim support that produces nonsense.
- Why only 20 pages at a time?
- Recognition takes a few seconds per page on our hardware, and a request has a time limit. Split a longer document first and run it in parts.
- Does it change how the page looks?
- No. The original image of every page is kept exactly; the recognised text is written in an invisible layer that only search and copy can see.
- Why does this tool need a Pro subscription?
- Because it runs on hardware we pay for. Everything that runs in your own browser stays free with no account, at any file size. The cloud tools are the few jobs a browser tab genuinely cannot do, and Pro is what pays for the machines that do them.
Also called Scanned PDF to text, Make a PDF searchable, Text recognition.