How it works
- Pick your video. Nothing is uploaded — the audio is read and transcribed on your phone or computer.
- Press Generate subtitles. Whisper, an open speech-recognition model, writes down every word with its timing.
- Tap any line to jump to it and correct a word. Timing stays in sync.
- Choose a style and download the MP4 with the captions burned in, or an .srt for YouTube, CapCut or Premiere.
Frequently asked questions
Is it really free?
Yes — no sign-up, no watermark, no limits on the number of videos.
Is my video uploaded?
No. Speech recognition and video rendering both run on your device. Only the speech model itself is downloaded once.
How accurate is the Arabic?
Clear Modern Standard Arabic comes out well; strong dialects and background music lower the accuracy, so review the text before exporting. “Best” accuracy helps.
How long does it take?
Roughly real time for transcription on a laptop with “Fast”, and about the length of the video to render. Phones are slower.