Images & scanned PDFs → editable text

Image to Text / OCR

Highly recommended

Extract printed text from photos, screenshots and scanned PDF pages in 16 languages. Crop visually, compare the text with the page, fix mistakes, then copy, download TXT or save a searchable PDF. Recognition runs in your browser; nothing is uploaded.

Many images & PDF pagesVisual cropUncertain words highlightedTXT & searchable PDF
  1. 1Add images or PDFs
  2. 2Crop & recognize
  3. 3Check & export

Add images or a scanned PDF

Drop files here or choose them. PNG, JPG and WebP images, and PDF pages (rendered as images).

Up to 50 pages in total · images up to 10 MB and 16 megapixels · PDFs up to 50 MB (first 50 pages). Printed text works best; handwriting is not supported.

Add an image or PDF to begin.

Using Image to Text / OCR

Extract printed text from photos, screenshots and scanned PDF pages in 16 languages. Read many pages at once, crop visually, check uncertain words and export TXT or a searchable PDF — all in your browser.

  1. Add images (PNG, JPG, WebP) or a scanned PDF — drop them on the page or choose them. Up to 50 pages; each PDF page becomes an image. You can also try the two-page example.
  2. Choose the language (and an optional second language) and a text layout. Automatic suits documents; One line suits labels; Scattered text suits signs.
  3. Optionally rotate a page, pick a clean-up such as High contrast or Light text on dark, and drag on the page to read only one area. Apply to all pages copies these settings.
  4. Press Recognize. The engine loads once and then reads each page in turn; a confidence badge appears next to every page.
  5. Check the orange boxes — words the engine was unsure about — and click one to find it in the text. Edit the text, then copy everything, download a TXT file or save a searchable PDF.

Example

Open a two-page scanned PDF, choose English and Automatic, then press Recognize all 2 pages.
Switch to All pages to see both pages with page headings, fix any word marked in orange, and download the TXT — or download a searchable PDF whose text you can select and search.

Questions & answers

Are my images or PDFs uploaded?

No. Images, PDF pages and the extracted text stay in your browser. Only the OCR engine and the language files you select download from this website the first time.

Can I convert a scanned PDF to text?

Yes. Add the PDF and each page (up to 50) is rendered as an image and read. If the PDF already contains selectable text, the tool tells you — PDF to Text extracts that text exactly.

What is a searchable PDF?

A PDF that shows your page images with an invisible text layer on top, so you can select, copy and search the text. Its hidden text comes from the recognition; edits you make in the text box are included in the TXT, not in the PDF layer.

Which languages are supported?

English, Simplified and Traditional Chinese, Hindi, Spanish, Arabic, French, Portuguese, Russian, German, Japanese, Turkish, Korean, Italian, Ukrainian and Hebrew. Choose a second language for mixed text; languages are not detected automatically.

What do the orange boxes and percentages mean?

The percentage is the engine's own confidence estimate for the page, and orange boxes mark individual words below 75%. They are hints for checking, not proof of errors — high-confidence text can still contain mistakes, especially names and numbers.

How can I get better results?

Use sharp, upright images, choose the right language, crop away unrelated areas and try High contrast for faint text or Light text on dark for white text. Very small text in a large page reads better when you crop that section.

Can it read handwriting or keep the layout?

It is designed for printed text. Handwriting, tables and exact formatting are not reconstructed; the output is plain text with page breaks.

Is there a limit?

Up to 50 pages per session, images up to 10 MB and 16 megapixels, and PDFs up to 50 MB. Each page is reduced to at most 4 megapixels for recognition to keep memory use manageable.

Guides for this tool

Examples and checks to help with your next step.

Help improve this tool

Report a problem or suggest an improvement

Describe the issue without pasting private tool input. Feedback goes to our admin inbox.

Find another tool · Read practical guides