How it works

Nothing added, nothing lost. Three models work on every recording before it reaches this page.

1. Restored by SAWT

Our restoration model rebuilds the voice from worn tapes, radio hiss, low-bitrate uploads and loud rooms. It is trained to keep every letter as recited: a Quran speech model checks the text before and after, so nothing is added, dropped or changed.

2. Every word timed

A Quran recognition model aligns the recitation to the Uthmani text, ayah by ayah and word by word, so the word being recited lights up as you listen. Tap any word to jump to it.

3. Then and now

The original and the restoration play together, locked to the same moment. Flip between them mid-ayah with one switch, or press O, and nothing skips.

Credits

Restoration: SAWT by Quran Lab. Audio restored with SAWT carries the credit "Restored with SAWT by Quran-Lab".
Translation: Saheeh International, via Tanzil.net.
Quran text: Uthmani script, Hafs from Asim. Typeface: KFGQPC Uthmanic Script Hafs.
Recordings: Belong to their reciters and original publishers. Every recording page links to its source.
Artwork: The scenes behind each recording were generated for this site and show no people. Reciter portraits are illustrations generated from real photos of each reciter, credited below.
Photos: Yasser al-Dosari: MAL MALDIVE, CC BY-SA 4.0; restored (cleaned and sharpened, the face unchanged) by Quran LabAbdul Basit Abdul Samad: Unknown author, Public domain (PD-South-Africa)Mohamed Refaat: Unknown, Public domain (PD-Egypt); restored (cleaned and sharpened, the face unchanged) by Quran LabMustafa Ismail: Unknown author, Public domain (PD-Egypt); restored (cleaned and sharpened, the face unchanged) by Quran Lab