Over 100 Languages
Recognition works for Latin, Cyrillic, Arabic, Chinese, Japanese, Korean and many other scripts, including vertical Chinese and Japanese text. Choosing the right language matters: it is the biggest single factor in accuracy.
Copy the text out of a screenshot or photo.
Paste a screenshot, drop a photo or open a scanned PDF, and copy the words out of it. This image to text converter uses OCR (optical character recognition) to read printed and typed text in more than 100 languages, then lets you copy it, save it as a TXT file or download a Word document. The image and the text appear side by side, so you can check every line.
JPG, PNG, WebP, HEIC, BMP, TIFF and scanned PDF up to 20 MB, up to 20 files at once; paste a screenshot, use your camera, or enter an image URL. Runs in your browser. Nothing is uploaded.
Add the image. Paste a screenshot with Ctrl+V or Cmd+V, drop files, take a photo with your camera, or enter an image URL.
Pick the language. English is used unless another is detected; choose from over 100, including Chinese, Japanese, Korean, Spanish, French, German, Russian and Arabic.
Let it read. Images are straightened and cleaned automatically before reading, which helps with phone photos and faint scans.
Check the result. Click any line of text to highlight where it came from in the image.
Copy or download as plain text, text with original line breaks, TXT or Word (.docx). With several images, the text can be combined into one file.
Recognition works for Latin, Cyrillic, Arabic, Chinese, Japanese, Korean and many other scripts, including vertical Chinese and Japanese text. Choosing the right language matters: it is the biggest single factor in accuracy.
There is no need to save a screenshot first. Take it, press paste on this page and the text appears. That is the fastest way to grab text from a video, a locked PDF, an app or an error dialog.
Before reading, the converter straightens tilted pages and boosts contrast between text and background. Small, low-resolution screenshots are enlarged first, because recognition works best when letters are large enough.
Keep the original line breaks for poems, addresses and code, or join lines into flowing paragraphs. Download the result as a TXT file or as a Word document ready to edit.
Add up to 20 images, such as photos of the pages of a document, and get the text of each one or of all of them in one file.
A project manager receives a screenshot of a long error message and a table of figures in a chat. Retyping it would take ten minutes and introduce mistakes. She pastes the screenshot here, copies the recognized text into the ticket, and checks the numbers against the highlighted lines in the image. The whole job takes under a minute, and the text can be searched later, unlike the screenshot.
A small law office has old typed letters that exist only on paper. A clerk photographs each page with a phone and adds all 15 photos in one batch. The converter straightens the pages, reads them, and combines the text into one Word document that can be searched and edited. Before filing the scans themselves, she combines the photos with image to PDF.
A university student photographs lecture slides and textbook pages to make notes. Instead of retyping quotes and definitions, she extracts text from each image and pastes it into her notes app. For slides in Spanish, she switches the language to Spanish so accents are read correctly, then translates the few passages she needs in a separate tool and keeps the original wording next to them.
A traveler in Japan photographs a printed menu and a museum sign. Selecting Japanese as the language turns the photos into text he can paste into a translation app to find out what to order and when the museum closes. Clear, printed signs read well; decorative fonts are less reliable, so he crops the photo to the plain text first with crop image for a better result.
OCR is very accurate on clear printed or typed text: screenshots, documents, books and labels usually come out nearly perfect. Accuracy drops with blurry photos, very small letters, busy backgrounds, curved pages, stylized fonts and, especially, handwriting, which this tool does not reliably recognize. For the best results, use a sharp, well-lit image, keep the text straight, and choose the correct language. If a photo is small or soft, enlarging it first with upscale image can help.
Looking for "image to text art", which turns a picture into letters and symbols? That is a different tool: try the ASCII art maker.
Yes. Text recognition runs in your own browser, so your images and the extracted text stay on your device. Nothing is uploaded, which matters for IDs, contracts, receipts and private messages. The first time you use a language, its recognition data is downloaded to your browser; your images are never sent anywhere.
Paste or drop the image, or enter an image URL, choose the language, and copy the text that appears. You can also download it as TXT or Word. Text is read in more than 100 languages, and the image is straightened and cleaned first.
Yes. Drop a JPG, PNG, WebP, HEIC, BMP or TIFF file on this page. To convert a JPG to a .txt file, choose TXT as the download format after the text appears. Files up to 20 MB work, up to 20 at once, and the text from several images can be combined into one file.
Yes. There is no sign-up, no login and no watermark. You can read up to 20 files of up to 20 MB each at once, in more than 100 languages, then copy the text or download it as TXT or Word. Recognition runs in your browser.
Take the screenshot and paste it straight onto this page with Ctrl+V or Cmd+V, then copy the recognized text. There is no need to save the screenshot first, which makes it the fastest way to grab text from a video, an app or an error dialog.
Open this page in your phone's browser, choose the photo or take a new one, and copy the text. On recent iPhones, Live Text also lets you press and hold text in a photo; if that does not appear, use this page in Safari for longer documents or batches. The same steps work on Android and Samsung browsers.
Choose Word (.docx) as the download format. The file opens in Microsoft Word, Google Docs or LibreOffice, ready to edit. Keep the original line breaks for addresses, poems and code, or join lines into flowing paragraphs before you download the document.
Not reliably. The OCR engine is built for printed and typed text. Neat block capitals may work, but cursive handwriting usually will not. For the best result with printed text, choose the right language, which is the biggest single factor in accuracy.
Yes. Each page of a scanned PDF is read like an image, and the text can be combined into one file. Scanned PDFs up to 20 MB are accepted, and tilted pages are straightened and contrast is boosted before reading, which helps with faint scans.
It uses OCR, a machine learning method trained to recognize characters. It reads the text that is in the image; it does not describe the picture or generate new wording. Click any line of the result to see where it came from in the image.
No. Recognition happens entirely in your browser, and your images stay on your device. That holds for screenshots, photos and scanned PDFs, and the text is only copied or downloaded where you choose. No account is needed, even for batches of 20.