Image to text converter

How to convert an image to text

1. Add your imageChoose a JPG, PNG, WebP, AVIF, BMP or GIF file, drop it on the box, or paste a screenshot with Ctrl+V or Cmd+V.
2. Pick the languageEnglish is selected by default. Tick another language if the picture contains it. You can combine up to three, for example English and Spanish.
3. Extract and editPress Extract text. When it finishes, fix any slips in the text box, then copy the text or download it as a .txt file.

What is image to text?

Image to text, also called OCR (optical character recognition), reads the letters in a picture and turns them into text you can select, search, edit and paste. It is the quick way to rescue words from a screenshot, a photographed page, a slide, a receipt or a sign without typing them again.

This tool uses the open-source Tesseract engine compiled to WebAssembly. The engine and its language files are served from this website and run on your own device. Your image is not sent to a server, which makes it a sensible choice for ID copies, invoices and private notes.

Get a more accurate result

Accuracy depends mostly on the picture. Printed text in a clear photo or a screenshot usually comes out almost perfectly. Blurry, tiny or low-contrast text produces mistakes, and no OCR tool can read what the camera did not capture.

• Use a sharp, well-lit photo and keep the page flat and straight.
• Aim for letters at least 20 pixels tall. Zoom in or crop to the text before you upload if needed.
• Dark text on a light background works best. Leave Image clean-up on Auto-contrast, or try Black and white for faded scans.
• Choose the right language. Reading German with the English model drops accents and splits words.
• Pick Single block for one paragraph, Single line for a caption, or Sparse text for labels scattered across a picture.

What it can and cannot read

It reads printed text in many fonts and handles columns reasonably well. Handwriting, decorative lettering, curved text and text on busy photographs are weak spots, so expect to correct those results by hand. Tables come back as lines of text rather than rows and columns.

The tool does not translate. If you are after image to text translate, extract the text here, then paste it into the translator you prefer. Because the text box is editable, you can also tidy the result before you copy it.

Languages and the first run

The first time you read an image, the page downloads the recognition engine and the language you chose, about 2 to 5 MB in total. After that your browser keeps the language file, so later runs start much faster. Available languages are English, Spanish, German, French, Portuguese, Italian and Japanese.

Images with a long side above 3000 pixels are scaled down before reading, which keeps the tool fast without hurting normal text.

Questions