Automatically assemble your images/videos to match a voiceover and script — completely free, runs entirely in your browser.
Drop your narration/voiceover file here (mp3, wav, m4a).
Leave one blank line between scenes — that's how scenes get split apart.
Auto-timing is just an estimate based on word count. "Auto-detect from audio" actually listens to the audio (an AI model that runs entirely in your browser, free) and works out much more accurate timing — the first run downloads a small model (~40-75MB), which then stays cached in your browser.
Upload as many files as you like, all at once. They're assigned to scenes in order by default — tap a scene, then tap a media item to reassign it.
Before exporting, preview it to make sure every scene's timing and media look right.
Export works best in Chrome or Edge (desktop). The file downloads in .webm format — use VLC or any free converter if you need .mp4. Recording takes as long as the audio itself (it's a real-time capture).