Photo2Text - home
Menu

Image to Text, Free Online OCR

Extract text from any image in seconds. 100% private: your images never leave your device.

Looking for script-specific OCR? Jump to Arabic OCR (العربية) or Urdu OCR (اردو او سی آر).

·

How it works

  1. 1. Upload, drop a JPG, PNG, or WebP, or take a photo on your phone.
  2. 2. Extract, pick a language and click Extract Text. The OCR model loads only then.
  3. 3. Copy, edit the result, copy it, or download a .txt file.

Powered by the open-source Tesseract OCR engine, running entirely in your browser via Tesseract.js.

Why private?

  • No uploads , image pixels stay on your device; we never see them.
  • No accounts, open the page and start; nothing to sign up for.
  • Works offline after first load , once the page and OCR models are cached by your browser, you can often reuse them without re-downloading (honest PWA-ish behavior, not a guaranteed installable app).

How to use Photo2Text, step by step

Using the tool takes less than a minute once you know the order. First, choose the image: a photo of a page, a screenshot, or a scan saved as JPG, PNG, or WebP. Keep the file under roughly 10MB, and use the original file rather than a copy that has already passed through a chat app, because each re-save throws away detail. Second, pick the language that matches the text. The homepage converter defaults to English, and the dedicated Arabic and Urdu pages load the right model for those scripts. Third, upload the file by dropping it onto the tool or tapping to choose it, and on a phone you can take a photo straight from the camera instead. Fourth, click Extract Text. The OCR model downloads on demand the first time, so the first run takes a few seconds longer, and later runs in the same session reuse it. Fifth, read the result against the image, fix any mistakes in the text box, then copy it or download it as a plain text file. That is the whole workflow: choose, upload, extract, check, copy. If the result looks wrong, the usual fix is a better image, not a different tool, so re-shoot or re-crop before you run it again.

Which page should you use for your file?

Photo2Text has one tool with several front doors, and picking the right door saves a step. Use the homepage converter for English text and for anything where the language is mixed or uncertain, because it is the general-purpose starting point. Use the Arabic OCR page when the page is in Arabic, and the Urdu OCR page when it is in Urdu, because each loads the matching language model and carries script-specific advice. Use screenshot to text when the source is a screen capture, an error dialog, or a chat message, since screenshots are sharp and the page is tuned for them. Use JPG to text for camera photos, which are almost always JPEG files, and PNG to text for crisp desktop captures and scans, since PNG keeps every pixel exactly as captured. If you are new to OCR, start with our step-by-step extraction guide, and if the text is Arabic, read the Arabic OCR guide first, because it covers fonts, diacritics, and right-to-left layout before you run your first extract.

Tips for the best results

Good OCR starts with a good image, and a few habits make a large difference. Photograph pages in soft, even light, with the phone parallel to the paper, so lines stay horizontal and no corner falls into shadow. Fill the frame with the paragraph you need instead of photographing a whole desk, and tap to focus on the middle of the text before you shoot. Crop away margins, stamps, seals, and neighbouring columns, because everything in the frame that is not your text is noise the engine has to sort through. When a page has two columns, crop one column at a time rather than running the whole page at once. Pick the language that matches most of the page, and use a combined mode only when two scripts genuinely share the same crop. Check the result against the image line by line, and pay special attention to names, dates, amounts, and phone numbers, because a single wrong digit changes the meaning completely. Keep the original image until the text is saved and checked, because re-running OCR on the same image is quick, but re-taking a photo of a page you have already returned is not. For a deeper version of these tips, read our Urdu OCR tips and the Arabic OCR guide.

What can you extract text from?

People use Photo2Text for everyday jobs that used to mean typing by hand. Students copy a paragraph from a textbook or a whiteboard into their notes. Office workers pull the text out of a scanned letter, a certificate, or a contract page so they can search it, quote it, or translate it. Shop owners read a receipt or a supplier invoice without retyping every line. Developers and support staff copy error messages from screenshots instead of typing them from memory. Readers of Arabic and Urdu extract a passage from a book or a newspaper clipping to share it or save it as text. The common thread is printed text in reasonable condition: a clear photo or scan of ordinary print converts well, while casual handwriting, decorative fonts, and very low resolution images do not. If your document is printed, flat, and readable when you zoom in, it is a good candidate. Try it on the homepage converter, and if the text is in Arabic, the Arabic OCR page is the faster route.

What happens to your text after extraction?

The text you extract is yours to keep and reuse however you like. Copy it into a document, paste it into a message or an email, search it to find the original source online, or drop it into a translation tool when the document is in a language you are still learning. Because the work happens in your browser, there is no upload queue to wait for and no account page to manage: you open the page, extract, and move on. If the document matters, save the text file next to the original image, so you can re-check any line later without photographing the page again.

OCR by language & use case

Start with Arabic OCR or Urdu OCR when your photo is in those scripts.

Also: JPG to Text · PNG to Text

Free image to text converter in your browser

Photo2Text is an online OCR tool built for people who need text out of a photo without sending that photo to a stranger's server. Whether you snapped a whiteboard, a receipt, or a page of Urdu or Arabic print, you select the language, run Extract, and edit the result locally. We call it best-effort OCR because accuracy always depends on lighting, focus, resolution, and script complexity - especially for Nastaliq Urdu and connected Arabic letterforms.

Specialized OCR pages

Prefer a dedicated workflow? Use the Urdu image to text tool for اردو pages, the Arabic OCR converter for العربية print, or pull text from a screenshot when the source is a UI capture. Camera files usually open best via JPG to text; crisp desktop captures belong on PNG to text.

New to OCR? Start with how to extract text from an image. For Nastaliq and crop tips, read Urdu OCR tips. Working with Arabic print? Read our Arabic OCR guide. Comparing options? See our honest free OCR tools comparison. More about the project on the About page, Contact, Privacy Policy, or Terms of Use. Browse every URL on the HTML sitemap.

Sources & further reading

Frequently asked questions

Yes, Photo2Text is free to use with no signup and no daily quotas on our side. OCR runs in your browser, so the only limits are your device speed and the size of the image you choose (about 10MB).