Skip to content
zspace tools

Image Tools

Image to Text (OCR)

Pull the text out of a photo, screenshot or scanned page. Choose the language, tidy the image if needed, then copy the text or download it as TXT or Word — all done on your device.

Runs in your browser

Features

What you get

  • Six languages

    English, Hindi, Gujarati, Spanish, French and German — combine up to three for mixed-language text.

  • Clean up first

    Rotate, crop to the text, greyscale, contrast and upscaling — on a copy, never your original.

  • Editable result

    Text appears in an editor next to the image so you can compare and correct it.

  • Confidence shown

    Overall confidence and the number of uncertain words tell you how carefully to check.

  • Copy, TXT or Word

    Copy the text, or download it as a .txt or .docx file.

  • Private

    Your image is read on your device. Nothing is uploaded.

How to use

How to use the Image to Text (OCR)

Tesseract's neural-network (LSTM) recogniser reads the image line by line using the language models you pick. Before reading, a copy of your image can be rotated, cropped to the text, converted to greyscale, given more contrast and enlarged — small text is upscaled automatically because OCR works best at around 300 DPI. The engine and models are served from this site and cached by your browser after the first use. You get an overall confidence score and a count of uncertain words, so you know when to check the result carefully.

  1. 01

    Add an image

    A photo, screenshot or scan in JPG, PNG, WebP, BMP or TIFF.

  2. 02

    Pick the language

    Choose the language(s) in the image. Rotate or crop if the text is sideways or surrounded by clutter.

  3. 03

    Extract and check

    Press Extract text, review the result next to the image, then copy or download it.

Examples

Worked examples

  • A screenshot

    A screenshot of an email reads almost perfectly — clean digital text usually scores over 90% confidence.

  • A Hindi notice

    Choose Hindi (add English if the notice mixes both) to read a printed Devanagari notice.

  • A phone photo of a page

    Rotate it upright, crop to the page and add a little contrast before extracting — accuracy improves noticeably.

Technical specifications

Engine
Tesseract 5 (LSTM) via tesseract.js
Languages
English, Hindi, Gujarati, Spanish, French, German
Input
JPG, PNG, WebP, BMP, GIF, TIFF — up to 25 MB
Output
Editable text, TXT, DOCX
First use
Downloads ~4 MB engine + 1–3 MB per language (cached)
Processing
In your browser — nothing is uploaded

Privacy

Images are read in your browser and never uploaded. The OCR engine and language data are downloaded from this site and cached by your browser.

Read the full privacy policy.

FAQ

Image to Text (OCR) questions

An image to text converter uses OCR (optical character recognition) to turn the words in a picture into text you can edit, search and copy. ZSpace Tools' Image to Text converter uses Tesseract, a long-established open-source OCR engine, running in your browser with language models for English, Hindi, Gujarati, Spanish, French and German.

Tesseract's neural-network (LSTM) recogniser reads the image line by line using the language models you pick. Before reading, a copy of your image can be rotated, cropped to the text, converted to greyscale, given more contrast and enlarged — small text is upscaled automatically because OCR works best at around 300 DPI. The engine and models are served from this site and cached by your browser after the first use. You get an overall confidence score and a count of uncertain words, so you know when to check the result carefully.

Very good for clear, printed text and screenshots; lower for blurry photos, low light, small text, unusual fonts and complex layouts. It's never 100% — always check important details like names, numbers and amounts.

Only poorly. Tesseract is trained on printed text, so handwriting is often misread.

Text in simple tables is read row by row, but the table structure isn't rebuilt. Complex tables may come out out of order.

English, Hindi, Gujarati, Spanish, French and German. Pick every language that appears in the image, up to three.

No. OCR runs in your browser. Only the OCR engine and language files are downloaded — from this site, not a third party.

For PDFs, use the PDF to Word Converter — it keeps the PDF's existing text and can OCR scanned pages.