Quran Whisper · Tajweed

Pick an ayah, recite it (mic or upload), and get lahn jaliy & lahn khafiy feedback — quantitative + qualitative.

1 · Reference (النص المرجعي)

or edit / paste your own text below ⤵
…
This text is what your recitation is scored against. Leave empty to transcribe only (no scoring).

2 · Your recitation (تسجيلك)

🎤 Record
⬆️ Upload
00:00
📁

Drag & drop audio here

or click to browse · mp3, wav, m4a, ogg, webm…

3 · Evaluate

Raw JSON response

    

  


  
  

    
Settings

    

      

        
        
      

      

        
        
      

    

    
Audio is POSTed to <API base>/transcribe as multipart form-data
      (fields: audio, expected, model).
      Change the model field to try a different recitation model with no redeploy. 

      ⚠️ Lahn khafiy here is a transcription-derived heuristic — true madd/ghunnah/qalqalah scoring needs an acoustic phoneme model.