Drop a video or an audio file and MAI-Reel writes down what is said, with a timestamp on every block. The speech model downloads once and runs on your device: the file is never uploaded.
Captions with real timings, word by word when the audio allows it.
Download .SRT for YouTube, TikTok or Premiere, or .TXT to read it.
Spanish, English, French, German, Italian and automatic detection.