Which OCR engine is this?

Tesseract, compiled to WebAssembly and running in your browser — the same engine behind our image-to-text tool. The recognition model is about 5 MB and is downloaded once, then cached, so the second document costs nothing extra.

What is a searchable PDF?

Your scan, unchanged, with the recognised words written over it in invisible text. The pages still look exactly as they did — we do not re-render them, so the image data, compression and bookmarks all survive — but Ctrl-F, copy and paste, and anything that extracts text will now find the words.

My PDF already has text. Should I use this?

No, and the tool will tell you so. If a page already carries text, that text is what the document actually says; recognition can only guess at the same words from their shapes and will guess some of them wrong. Existing text is used as-is, and a fully searchable document is sent to the Markdown converter instead.

How good is the result?

It depends almost entirely on the scan. Clean, straight, 300-dpi black text on white does very well; a phone photo of a curled receipt does not. There is no image clean-up step here — no deskewing or contrast correction — so what you feed it is what it reads. The mean confidence and any empty pages are reported so you can judge rather than guess.

Does it handle languages other than English?

Not yet. The recognition model shipped here is the English one. A document in another language will produce something, and that something will be wrong in ways that are hard to spot, so treat English as the boundary rather than a preference.

Are my PDFs uploaded to a server?

No. Every operation runs entirely in your browser on your own device. Your files are never uploaded anywhere, and nothing is stored once you close the tab.

Is it really free? Are there limits?

Yes, it is completely free with no account required. There are no page or file-count limits and no watermarks — because the work happens on your device, there is nothing for us to meter.

OCR a scanned PDF

Read the words out of a PDF that is really a picture of a page. You get the text, and a searchable copy of the original with an invisible text layer over the scan — so it looks exactly the same and Ctrl-F finds things. Your files are never uploaded — everything runs on your device.

Your PDF
Drop a scanned PDF here
PDF
Your files are never uploaded — everything runs on your device.
Before you start
Text

Free & unlimited — if it helped, buy me a coffee

How to OCR a PDF

Open your scanned PDF

Drag and drop it, or click to choose it. The document is read here in the browser — nothing is uploaded, which matters more than usual for the kind of paperwork people scan.
Check what it will cost

Before anything runs you get the page count, the one-time model download, and a time range. Pages that already contain text are skipped, and if every page has text the tool says so instead of running.
Read & download

Take the plain text, or the searchable PDF. It also reports pages that came back empty rather than quietly leaving holes in the result.

What you can do

Merge & split

Combine several PDFs into one, or pull out just the pages you need and download them as a separate file.

PNG to PDF, JPG to PDF

Drop in PNG or JPG images and each one becomes a PDF page you can arrange alongside your documents, then download as a single PDF.

PDF to PNG, PDF to JPG

Go the other way too: render each page (or just the ones you select) to PNG or JPG and download them together as a ZIP.

Rearrange PDF pages

Drag a page by its handle to move it. Works with a mouse and with touch, so you can reorder pages on your phone too.

Rotate & auto-orient

Turn any page 90° at a time, or let Auto-rotate detect sideways and upside-down scans and straighten them for you.

Preview any page

Click a page to open a crisp full-size preview and step through the document before you download.

Delete & extract

Remove pages you do not need, or extract a range by selecting pages and downloading only the selection.

Reduce the file size

Make a PDF smaller: a lossless re-save that keeps selectable text, or a stronger image setting for scans and image-heavy files.

Private and free

No accounts, no page limits, no watermarks. Your files are never uploaded — every operation runs locally in your browser.

Frequently asked questions

Which OCR engine is this?: Tesseract, compiled to WebAssembly and running in your browser — the same engine behind our image-to-text tool. The recognition model is about 5 MB and is downloaded once, then cached, so the second document costs nothing extra.
What is a searchable PDF?: Your scan, unchanged, with the recognised words written over it in invisible text. The pages still look exactly as they did — we do not re-render them, so the image data, compression and bookmarks all survive — but Ctrl-F, copy and paste, and anything that extracts text will now find the words.
My PDF already has text. Should I use this?: No, and the tool will tell you so. If a page already carries text, that text is what the document actually says; recognition can only guess at the same words from their shapes and will guess some of them wrong. Existing text is used as-is, and a fully searchable document is sent to the Markdown converter instead.
How good is the result?: It depends almost entirely on the scan. Clean, straight, 300-dpi black text on white does very well; a phone photo of a curled receipt does not. There is no image clean-up step here — no deskewing or contrast correction — so what you feed it is what it reads. The mean confidence and any empty pages are reported so you can judge rather than guess.
Does it handle languages other than English?: Not yet. The recognition model shipped here is the English one. A document in another language will produce something, and that something will be wrong in ways that are hard to spot, so treat English as the boundary rather than a preference.
Are my PDFs uploaded to a server?: No. Every operation runs entirely in your browser on your own device. Your files are never uploaded anywhere, and nothing is stored once you close the tab.
Is it really free? Are there limits?: Yes, it is completely free with no account required. There are no page or file-count limits and no watermarks — because the work happens on your device, there is nothing for us to meter.

Your PDF

Before you start

Text