SpeakCove
Get the app
Web and app

Scans and paper, read aloud.

A scanned page is a picture of letters, so SpeakCove recognizes the text first and then reads it to you. On this website that works for scanned PDFs; in the app, you point the camera at paper.

Fig. 1 · Scanning

From a picture of a page to a voice.

(a) A PDF without a text layer goes through text recognition in your browser, line by line, and is then read aloud a sentence at a time. (b) In the app for iPhone and iPad, the camera scans the paper and the text is recognized on the device.

How it works2
  1. Bring the scan.

    WebPick the language of the text, then open the scanned PDF.AppTap + and Scan document. Photograph the pages, put them in order, then save.

  2. The text is recognized.

    WebEvery page goes through text recognition in your browser: Recognizing text… page 3 of 12AppYour phone or tablet recognizes the pages on the device. Nothing is uploaded.

  3. Press play.

    The voice reads the recognized text with the spoken sentence highlighted on the page. Tap any sentence to go there.

Notes3

Scans, then speech

What happens to a scanned page.

Many PDFs are not text at all. A chapter run through the office copier, an article saved from an old journal, a letter someone photographed: each page is a picture of letters, and an ordinary reader finds nothing in it to read. SpeakCove recognizes the letters first, then reads them aloud like any other page.

On the web this happens on its own. When you open a PDF and none of its pages carries text, the reader runs text recognition right in the page, one page after another, and counts them as it goes. Then press play. The highlight moves over the scanned page itself, because every recognized line is laid back over the picture where it was found, and you can tap a line to jump there.

Recognition reads in the language you pick, so set it before you open the file; change it later and the pages are recognized again. All 31 languages work. Two limits are worth knowing. Every page is recognized before reading starts, so a long scan takes a while. And the reader only recognizes PDFs that are scanned throughout: if some pages already carry text, it reads those and skips the ones that are only pictures.

Paper on your desk is a job for the app, on iPhone and iPad. Tap + and Scan document, and the camera finds the edges of the page. Scan as many pages as you need, turn any that came out sideways, then save them to your library as a PDF. The phone recognizes the text itself, with Apple’s text recognition on iPhone and iPad, so nothing is uploaded. On iPhone and iPad the app picks the voice language from the text it finds. The first pages are recognized when you save, the rest while it reads.

The app also checks each page on its own. A PDF that mixes typed and scanned pages reads from start to finish, and a page whose text comes out as garbled symbols is recognized from the picture instead.

Details4
Where
WebThe free reader on this website, in your browserAppSpeakCove for iPhone and iPad (iOS 26 or later)
What it reads
WebPDFs whose pages have no text layerAppPrinted paper through the camera, several pages per scan
Recognition
WebPaddleOCR, running in your browser tabAppOn the device: Apple’s text recognition on iPhone and iPad
Languages
WebAll 31, in the language you pickChange the language and the pages are recognized again.AppiPhone and iPad: 22 of 31, detected from the textEnglish, Arabic, Czech, Danish, Dutch, Finnish, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese
Progress
Web“Recognizing text… page 3 of 12”AppShown on the document in your library
Reading starts
WebOnce every page is recognizedAppAfter the first pages; the rest are recognized while it reads
Mixed PDFs
WebReads the pages with text, skips picture-only pagesAppChecks each page; recognizes pictures and garbled text
Before saving
WebNothing to set upAppReorder, rotate or delete pages, or scan more
Highlight
WebThe spoken sentence, on the scanned page; tap to jumpAppThe spoken sentence, on the page; tap to jump
Upload
WebNone. Only the engine and language data download.AppNone
Questions5
Can SpeakCove read a scanned PDF aloud?

Yes. Open it in the reader on this website. If none of its pages has text, the reader recognizes the text in your browser first, then reads it with the spoken sentence highlighted.

Can I take a photo of a page and hear it?

In the app, yes: tap + and Scan document and point the camera at the page. The web reader takes files, so a photo needs to be a PDF first.

Is my scan uploaded anywhere?

No. On the web, recognition runs in your browser tab and only the recognition engine and language data download. In the app it runs on your phone or tablet.

What if it finds no text?

Very faint or blurred scans can come back empty. The web reader then says “No readable text found in this PDF.” Try a sharper scan, or scan the paper with the app.

Try it6

Have a scanned PDF?

Open it in the free reader. No sign-up, nothing to install. For paper on your desk, scan it with the app.

More features7
Web and app · Formats

File formats

Ten formats, from PDF and EPUB to comics, in the browser and in the app.

Web and app · Languages

31 languages

Every voice reads all 31 languages, from English and Español to 日本語.

Web and app · Privacy

Privacy

No account, no install, and your file is never uploaded.

Related Posts
All posts →