Automatic captions with on-device AI
Add captions to any video and burn them into the picture. Generate them automatically with AI speech recognition that runs on your own device, import an SRT or VTT file, paste a transcript to auto-time it, or type captions while the video plays. Style them with presets such as YouTube, TikTok bold, karaoke and lower third, including word-by-word highlighting, then export an MP4 with open captions plus SRT/VTT files.
Click Auto captions (AI) and the speech in your video is transcribed by Whisper, an open speech-recognition model, running inside your browser. The model is downloaded once (about 41 MB for Fast, 77 MB for Accurate) and cached; your audio is never uploaded. It supports 17+ languages, records the timing of every word for perfectly synced karaoke highlighting, and splits the text into short, readable captions you can then correct.


