Image to text
Turn photos of documents, scans, screenshots or scanned PDFs into editable text – with the Tesseract text recognition (OCR) right in your browser. Nothing is uploaded, and this page does not even need any third-party service for it: everything comes from mytoolio itself. 20 languages, several pages at once, export as text or searchable PDF.
Runs entirely on files from mytoolio.org – no third parties.
Nothing recognised yet. Drop a photo, scan, screenshot or PDF here.
How it works
Choose a file
Drop photos, screenshots or a PDF into the box, pick them or paste an image with Ctrl+V. Several pages are recognised one after another.
Pick languages
Tick every language that appears in the text. PDFs with real text are read directly, scanned PDFs via text recognition.
Check & use
Uncertain words are highlighted in orange on the image. Correct the text in the box and copy it, save it as TXT or turn all pages into a searchable PDF.
Frequently asked questions
Are my images uploaded?+
No. Images and PDFs are only read in your browser and recognised there. The text recognition (Tesseract), its language data and the PDF reader come from mytoolio.org itself – there are no requests to other services, and nothing from your files leaves your device.
How accurate is the recognition?+
Very accurate for good-quality printed text – usually well over 95% of the words are right. Hard cases are handwriting, very small or distorted type, pages photographed at an angle, low contrast and unusual fonts. Tip: photograph the page straight and well lit; pages shot at an angle can be fixed with “Straighten”, and multi-column pages and tables read better with the matching page layout. The tool highlights uncertain words in orange.
Can I get the text out of PDFs?+
Yes. If the PDF already contains real text (e.g. exported from Word), it is taken over directly and without errors. Scanned PDFs that only contain images of the pages are read page by page with text recognition. “Ignore PDF text” always forces text recognition.
Which languages are supported?+
20 languages: German, English, French, Spanish, Italian, Dutch, Polish, Portuguese, Turkish, Czech, Danish, Swedish, Norwegian, Finnish, Hungarian, Romanian, Croatian, Russian, Ukrainian and Greek – mixed within one text, too. Pick only the languages that actually appear: that is faster and more accurate.
Does it recognise handwriting?+
Only to a very limited degree. Tesseract is made for printed text: neat block letters sometimes work, joined-up handwriting usually does not.
What does the confidence percentage mean?+
It is the average confidence of the text recognition for the recognised words. Words below 60% are highlighted in orange on the image – worth a closer look.
How do I create a searchable PDF?+
Click “Searchable PDF”: every page is kept as an image with the recognised text invisibly on top. You can then search the PDF and select and copy text – it looks just like the original. Pages that already contained real text are taken over unchanged. Corrections in the text box do not change the PDF; it contains the recognition result.
Text recognition with Tesseract.js (Apache 2.0) and the “tessdata best” language data, PDF handling with pdf.js (Apache 2.0) – all served by mytoolio.org and run entirely in your browser. Results may contain mistakes; please check before sharing.