Extract text from an image

What is OCR?

OCR stands for optical character recognition. It reads the text in a picture, such as a screenshot, a scanned page or a photo of a document. You get plain characters back that you can search, copy and edit.

How to extract text from an image

Choose the language of the text, then drop a PNG, JPG, WebP, BMP or GIF image above. Tesseract.js, a WebAssembly build of the Tesseract OCR engine, reads the image in your browser while a progress bar shows how far it has got. The recognized text appears in a box below the image. Copy it with one click or download it as a .txt file. Run another image with a different language whenever you need to.

Why run OCR in your browser?

Online OCR services ask you to upload your documents, contents and all. Here the Tesseract engine runs on your own device. The image never leaves your computer.

Tips for better accuracy

Use the sharpest image you have, with strong contrast. Pick the right language and crop out busy backgrounds. Clean scans and screenshots read much better than tilted phone photos. Small text reads better when the image is bigger, so upscale a tiny screenshot first. Handwriting and stylized fonts are read poorly by Tesseract, and tables come out as lines of text without their grid.

Before you start

Input: PNG, JPG, JPEG, WEBP, BMP, and GIF. Output: Text. The settings appear before anything runs, so you can check them first. Your original file is never changed.

Local AI tools may download a model before the first run and can use a lot of memory. The output can contain mistakes or gaps, so review it before you rely on it.

Keep the original until you have opened and checked the result. The work happens in this browser tab, so the size you can handle depends on your device's memory, your browser and the file itself.

Steps to extract text from an image

1. Choose the language of the text in your image.
2. Drop an image or screenshot into the box above.
3. Wait while the OCR model loads and reads the image. A progress bar shows how far along it is.
4. Copy the text or download it as a .txt file.