Give your audio and video a new dimension.
Saved in this browser. Download a project backup before clearing browser data.
Import media, or create an ambient pad.
Processed preview uses 360p. Update it after edits to hear and see the current result.
0:00
Keep this tab open. Your media is not uploaded.
Portrait swaps width and height; square uses the shorter edge. Larger output does not restore missing source detail. Try a 5-second section first.
1. Choose a card and import your media.
2. Edit captions, sound, visuals, and pad progressions in the shared editor.
3. Update the processed preview, listen, then export.
Media stays on your device. English auto-captioning downloads an open speech model from Hugging Face (roughly 40–80 MB initially). Only model files are requested; audio is processed locally.
Start with short projects (under 2 minutes recommended). Higher resolutions and longer files may exceed your browser's memory. MOV playback depends on its codecs; HEVC is not guaranteed. Working exports use 24 fps and mono audio at 24 kHz in this first release.
Supported now: cuts/fade-to-black, one image overlay, spectral noise reduction, editable captions and word-timed karaoke, layered pads, processed previews, MP4/WebM/WAV export, local projects and media-inclusive backups.
Still to come: true overlapping video crossfades, multi-video picture-in-picture, blurred fill, per-caption style overrides, draggable multi-track timing, project style presets, stereo processing, and additional caption languages. No MP3 export yet.
Saving is browser-local and can fail if storage is full. Download a backup. Microphone recording requires HTTPS and permission. Do not close this tab during processing.