Image to text: copy the text from a picture
To copy the text from a picture, add the photo, screenshot or scan here: its text is read on your device with Tesseract OCR and shown ready to copy, edit or save as a .txt file. It reads printed and typed text in English and 18 more languages, from JPG, PNG, iPhone HEIC photos and scanned PDFs, and the picture is never uploaded. Handwriting, tiny print and blurry photos are read poorly.
Drop a picture with text hereChoose a picture with text
A photo, a screenshot or a scanned PDF. The text is read on your device.
Your files stay on this device. Nothing is uploaded.
How to image to text: copy the text from a picture
Add the picture
Add a photo, screenshot or PDF on tuul.io/image-to-text/: JPG, PNG, WebP, iPhone HEIC or PDF, one or many. The language of the text starts as your browser's; pick another first if the text is in another language.
Wait while it's read
The first time, the text reader and the language file download (about 7 MB with English; the row shows how far along it is); after that a page takes a few seconds, and a PDF's pages appear as each is read. Photos are turned upright by their camera setting, and a sideways scan is tried turned.
Copy or save the text
The text appears under the drop zone. Fix any mistakes in the box, then press Copy text, or Download TXT for a text file. Several files download together as a ZIP.
Image to text: details
- Input
- JPG, PNG, WebP, HEIC (iPhone), AVIF, GIF, BMP and PDF (every page; a page with its own text gives that text exactly, and only pages without text are read as pictures)
- Languages
- 19: English, Arabic, Chinese, simplified, Chinese, traditional, Dutch, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Turkish, Ukrainian, Vietnamese
- Output
- Text to copy or edit, or a .txt file (UTF-8) for each picture; a ZIP for many
- Accuracy
- Word for word on screenshots and clean printed pages in our tests (English, Turkish, German, Spanish, French and Russian, 9-pixel text included); about 6% of characters wrong on a phone photo of a page
- Speed
- After the first download, 1 to 3 seconds for a screenshot or a page, 4 seconds for a two-page scanned PDF, about 6 for a sideways page (it's tried turned), on an ordinary laptop
- Reads well
- Printed and typed text: screenshots, documents, receipts, signs, book pages
- Reads poorly
- Handwriting, text under about 10 pixels tall, blurry or steeply angled photos, decorative fonts; tables come out as lines of text
- Download
- About 7 MB the first time: the text reader (3.9 MB) and the English file (3 MB), or another language's (0.7 to 3 MB), from jsDelivr, kept by the browser for next time; your picture is never sent
- Engine
- Tesseract OCR 5 (LSTM) through tesseract.js, both Apache 2.0
- Price
- Free, no sign-up, no limit
Image to text: questions
How do I copy text from a picture?
Open tuul.io/image-to-text/, add the picture and wait a few seconds while the text is read on your device. The text appears in a box: press Copy text and paste it anywhere, or Download TXT to save it as a file.
Is my picture uploaded?
No. Image to text reads the picture in your browser. The only downloads are the text reader and the language file, the first time; no request carries your picture or its name.
Can I copy text from a picture on an iPhone or Android phone?
Yes. On an iPhone XS or later with iOS 15 or later, open the photo in Photos, touch and hold a word (or tap the Live Text button) and tap Copy. On Android, open the photo in Google Photos, tap Google Lens, then select the words and copy them. tuul.io/image-to-text/ also works in Safari and Chrome on a phone, for many pictures at once or a whole scanned PDF.
Can it read handwriting?
Not well. The reader is trained on printed and typed text, so neat block capitals may come out partly right and joined-up handwriting mostly wrong. For handwriting, Live Text on an iPhone or Google Lens on Android does better.
Can I get the text out of a scanned PDF?
Yes. Add the PDF on tuul.io/image-to-text/: pages that have their own text (an invoice, a bank statement, an e-ticket) give it exactly and at once, and pages without text, such as scans, are drawn at 300 dpi and read. The text comes out page by page, and each page shows as soon as it's read; the row says which pages were read as pictures.
Which languages does it read?
19: English, Arabic, Chinese, simplified, Chinese, traditional, Dutch, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Turkish, Ukrainian, Vietnamese. The language starts as your browser's when it's one of these; choose another before adding the picture, and each language's file downloads once. English is read with the English file, which also gets most other Latin-alphabet text, but without letters like ğ, ß or ñ.
Why is the text full of mistakes?
Usually the picture: text too small, blurry, at an angle, on a busy background, or in a decorative font. Take the photo straight on, close enough that the letters are clear, and choose the right language. The note under the text says how sure the reading is.
How do I copy text from a picture on Windows or a Mac?
On a Mac, Preview and Photos let you select text in a picture with Live Text (macOS Monterey or later). On Windows 11, the Snipping Tool's Text actions button copies text from a screenshot. For a saved picture, a batch, or a scanned PDF, use tuul.io/image-to-text/ in any browser.
Sources
- Tesseract OCR (Apache 2.0) (github.com)
- tesseract.js (Apache 2.0) (github.com)
- Apple Support: Copy and translate text from photos on your iPhone or iPad (support.apple.com)
- Google Photos Help: Get info about your photos and surroundings (Google Lens) (support.google.com)
Next steps
- Crop imageCut a photo to the part you want: drag a box, pick 1:1, 4:5 or 16:9, or type exact pixels.
- Rotate imageNewTurn sideways photos the right way up, or mirror them, many at once.
- Compress imagesMake photos smaller in KB or MB, many at once: by quality, or to a size like 100 KB.
- Images to PDFPut photos and scans into one PDF, in the order you choose.
- Kindle highlightsNewExport My Clippings.txt as Markdown, CSV or PDF, book by book.
Guides
- How to copy text from a pictureCopy the text from a photo, screenshot or scanned PDF on an iPhone, Android phone, Windows or Mac, or in any browser with no upload.
More image tools
- HEIC to JPGTurn iPhone photos (HEIC) into JPG files, many at once.
- Resize imagesChange the size in pixels, many at once: exactly 1080 × 1080, half the size, or the longest side.Instagram photo sizeNewYouTube thumbnail sizeNew
- Remove location from photosTake the GPS location, EXIF and other hidden metadata out of photos, losslessly.
- Add text to a photoWrite a caption or meme text on your own photo: font, size, colour, outline, drag to place.
- GIF makerPut photos or images into an animated GIF, in your order and at your speed.
- Compress GIFMake animated GIFs smaller: fewer colours, fewer frames or a smaller size, timing kept.
- Add a watermarkNewPut your name or logo on many photos at once: a corner, the centre or tiled, see-through.
- Extract images from WordNewSave every picture in a Word, PowerPoint or Excel file, in its original quality.
- Motion photo to MP4NewSave the clip inside Samsung and Pixel motion photos as MP4, or keep only the photo.
- Extract images from PowerPointNewSave every picture, video and sound in a PowerPoint, named by slide.
- Convert imagesBetween JPG, PNG and WebP, and from HEIC, AVIF, BMP, GIF and SVG, many at once.To JPGWebP to JPGPNG to JPGAVIF to JPGGIF to JPGBMP to JPGTo PNGWebP to PNGHEIC to PNGJPG to PNGSVG to PNGAVIF to PNGTo WebPJPG to WebPPNG to WebP
- Extract images from ExcelNewSave every picture in an Excel workbook, named after the cell it sits on.