To extract text from an image, open the Image to Text tool, choose your picture or PDF, pick the language of the text and click extract. The recognized words appear on the page, ready to copy or download as a .txt file.
The tool reads printed text from JPG, PNG, WebP, GIF, BMP and TIFF images and from PDFs. Unlike most tools on this site, it sends the file to an OCR service, which is explained below.
On this page
Key takeaways
- Supported inputs are JPG, PNG, WebP, GIF, BMP, TIFF and PDF files.
- Choosing the right language improves accents and special characters.
- Images over about 1 MB are scaled down in your browser before they are read.
- Your file is sent to the OCR.space service for recognition, and it is not stored on our servers.
How to extract text in three steps
- 1
Choose your file. Open Image to Text and pick a JPG, PNG, WebP, GIF, BMP or TIFF image, or a PDF.
- 2
Select the language. The list covers 24 languages, including English, French, Spanish, German, Arabic, Chinese, Japanese and Russian. English is the default.
- 3
Extract and copy. Click extract. The text appears with a word count. Use Copy, or download it as a .txt file named after your image.
What it reads well
- Screenshots of articles, chats and error messages.
- Photos of printed pages, signs, receipts and book passages.
- Scanned documents saved as images or PDFs.
- Slides and posters with clear, large lettering.
Handwriting, decorative fonts and very small or blurred text are unreliable. For tips that raise accuracy, see how to improve Image to Text results.
File size and PDFs
The recognition service accepts files up to about 1 MB. If your picture is larger, the tool scales it down in your browser before sending it, which usually keeps the text readable. A long scanned PDF may be too big, so use Extract Pages to send a few pages at a time.
How your file is handled
Reading text from a picture needs recognition software that does not run in the browser here. Your file goes to the OCR.space service through this site's server and the text comes back. It is not stored on our servers. Avoid uploading documents that contain information you would not send to an outside service. Is Image to Text private? has the details.
What to do with the text
Paste it into a document, a spreadsheet or a translation tool. If the source was a PDF with real text, PDF to Text is more accurate because it reads the text directly, and PDF to Word gives you an editable DOCX.
Image to Text: FAQ
How do I extract text from an image?
Choose the image in Image to Text, select the language and click extract. Then copy the text or download it.
Which file types are supported?
JPG, PNG, WebP, GIF, BMP, TIFF and PDF.
Does it work on scanned PDFs?
Yes, for files within the size limit. Process long scans a few pages at a time.
Is my file uploaded?
Yes. It is sent to the OCR.space service for recognition and is not stored on our servers.
Which languages are available?
24, including English, French, Spanish, German, Italian, Portuguese, Arabic, Chinese, Japanese, Korean and Russian.
Why is the text not accurate?
Blurry, small or low-contrast images are hard to read. Try a sharper picture and choose the correct language.
Can it read handwriting?
Not reliably. It is built for printed text.
Is there a file size limit?
About 1 MB. Larger images are scaled down in your browser first.
Can I save the result?
Yes. Use Copy, or download the text as a .txt file.
From picture to copyable text
Choose the file, set the language and extract. Check the result against the original before you rely on it.
