Image Scanning: Listen to Photos, Screenshots, and Documents
Readox can pull text out of images and read it aloud, entirely on your device.
Not everything worth listening to lives on a web page. Textbook photos. Scanned documents. Screenshots of articles. Whiteboard notes. Until now, that text was stuck in pixels. Readox can now pull it out and read it aloud.
Everything runs locally on your device. No cloud servers, no uploads, no per-character fees.
How to scan an image
Three ways to get an image into Readox:
- Drag and drop an image file onto the sidepanel
- Paste a screenshot with Ctrl/Cmd+V
- File picker via the + button in the command bar, then “Open file”
Supported formats: JPEG, PNG, WebP, and SVG (up to 10 MB).
By default, extracted text starts playing immediately. You can turn off auto-play in Settings if you prefer to review the text first.
What happens when you scan something
First, Readox figures out what kind of content it’s looking at. Paragraphs. Titles. Tables. Formulas. Code blocks. Figures. Headers and footers. That matters because it helps Readox read the useful parts and skip the junk.
Then it reads the actual text. If something isn’t meant to be read straight through, like a figure or decorative element, it gets skipped instead of turning into nonsense audio.
What gets extracted
| Region type | What Readox does |
|---|---|
| Paragraphs, titles, lists, captions | Extracts text and reads it aloud |
| Tables | Marked in the transcript so you know they’re there |
| Formulas and equations | Skips with a “[formula]” marker |
| Algorithms and code | Skips with an “[algorithm]” marker |
| Figures, images, icons | Skipped silently |
| Headers and footers | Skipped silently |
Markers appear in the transcript so you know something was there. You just won’t hear garbled text from a chart or equation.
Best results with documents
This works best on structured documents: academic papers, research reports, books, magazines, contracts, and exams. That’s where it really shines. It can separate the main text from figures, tables, and math so listening feels much cleaner.
For these documents, you get:
- Body text extracted in reading order
- Tables, formulas, and figures identified and skipped
- Headers, footers, and page numbers filtered out
- Multiple columns handled correctly
Works with everything else too
For images that aren’t structured documents (UI screenshots, photos of signs, whiteboard notes, product labels), Readox just reads the image directly. You still get text extraction. Just without the extra document-aware cleanup.
More languages
English ships by default. You can also add language packs for:
- Latin: French, German, Spanish, Italian, Portuguese, and 27 more
- Cyrillic: Russian, Ukrainian, Bulgarian, Belarusian
- Korean
- Japanese
- Chinese: Simplified Chinese, Traditional Chinese
- Thai
- Greek
Language packs are quick one-time downloads you can install from Settings. In auto-detect mode, Readox tries all installed languages and picks the best match for each text line.
Free for everyone
Image scanning is free. No Pro subscription required. A quick one-time setup downloads what you need, and everything runs entirely on your device after that. No data ever leaves your browser.
And if something sounds off, send feedback. We’re always trying to make this better.
The whole idea behind Readox is simple. If there’s text in something, you should be able to hear it.
Read aloud web pages and PDFs with premium English voices that run on your device.