Image Tools

Image to Text (OCR)

Copy the text from a photo, screenshot, scan or PDF. Pick up to 3 languages, then copy the text or download it as a .txt file.

Drop an image or PDF here

or browse, or paste a screenshot with Ctrl+V

PNG, JPG, WebP, GIF, BMP, TIFF, HEIC up to 20 MB · PDF up to 50 MB (first 20 pages)

Languages in the image

Pick up to 3. Fewer languages give better results.

About Image to Text

Image to Text reads the words in a picture and gives them back as plain, editable text. It uses Tesseract, a widely used open-source OCR (optical character recognition) engine, and works with photos, screenshots, scanned pages and PDFs.

For the best result, choose only the languages that actually appear in the image, use a sharp, well-lit picture where the text is at least a few millimetres tall, and crop away anything that is not text. Printed text works far better than handwriting.

Need a searchable PDF instead of plain text? Use OCR PDF. To pull fields or tables out of documents, try the OCR layout extractor.

Frequently Asked Questions

Drop the image above (or paste a screenshot with Ctrl+V), tick the language of the text and click Extract text. When it is done, click Copy text or download the result as a .txt file.

English, Hindi, Spanish, Portuguese, Indonesian, German, French, Italian, Russian, Arabic, Chinese (Simplified), Japanese, Korean, Turkish, Dutch and Polish. Pick up to three at once for images that mix languages; choosing only the languages that appear gives the best accuracy.

Yes. Upload a PDF (up to 50 MB) or a multi-page TIFF and the text of up to 20 pages is extracted, with a marker before each page. For a searchable copy of the PDF itself, use OCR PDF instead.

Only with very neat block letters. The OCR engine is trained on printed text, so typed documents, screenshots, signs and book pages work far better than handwritten notes.

Tick Highlight uncertain lines to see how sure the OCR engine is. Amber lines may contain mistakes and red lines are probably wrong, so check them against the image before you use the text.

Use a sharp, evenly lit image where the text is large enough to read comfortably, crop away pictures and borders, and photograph pages straight on. Sideways or upside-down pages are turned upright automatically.

No. Files are uploaded over HTTPS, read, and deleted automatically after processing. The extracted text is not saved on the server.