Image to Text: Free OCR in Your Browser

Free image to text converter (OCR) that runs in your browser

Drop a photo, screenshot, scan or PDF on the box above, or paste a screenshot with Ctrl/⌘ + V. The text in it is read by optical character recognition (OCR) right in your browser and appears in a box you can edit, copy or save as a .txt file. Each line the reader found is boxed on the picture, coloured by how sure it is. If you want, translate the text into another language below.

What it does

• Private: the picture is never uploaded. The text reader (Tesseract) is downloaded into your browser and runs there.
• About 80 languages and scripts: English, Spanish, French, German, Russian, Arabic, Hebrew, Hindi, Chinese, Japanese, Korean, Thai, Vietnamese and many more. Choose the language of the text; its reading data is downloaded once and then cached.
• Phone photos come out the right way up (the camera's rotation is read from the file), huge photos are shrunk and small screenshots enlarged, so the reader sees letters at a good size.
• PDFs: the first five pages of a PDF are drawn and read, so scanned PDFs work too.
• Pick what you need: click line or word boxes to select them, then translate just those.

Questions

Is the image uploaded?

No. The picture is read in your browser. Only if you translate with the online engine is the text (never the picture) sent to the MyMemory translation service; with Chrome or Edge on a computer the translation also happens on your device.

Why are some words wrong?

OCR guesses letters from their shapes. Blurry, tilted or very small text, handwriting, decorative fonts and text over pictures are hard. Red and amber boxes mark the guesses it was unsure about. Try a sharper photo, crop closer, or pick the right language.

Does it read handwriting?

Neat block capitals sometimes, joined-up handwriting rarely. It is made for printed text.

How long does it take?

The first time, the reader and the language data are downloaded (a few MB). After that a page of text usually takes a few seconds, longer on phones and for big PDFs.

Can I get the text of a PDF without OCR?

If you can select the text in your PDF viewer, copying it is more accurate than OCR. This tool is for PDFs that are really pictures, such as scans.

Tips

• Hold the phone straight above the page, in good light, and fill the frame with the text.
• Tick Also read English words for signs or menus that mix a language with English words or numbers; untick it for a cleaner result in one language.
• For columns, newspapers or forms, select the boxes of one column and translate those.

Mostly translating typed text? The MES Translate page puts the text box first.