From source media to an exported AI video

1. Add media

Import images, video and audio from your device. Drag library items to their matching timeline tracks. Uploaded video can expose its source audio on a dedicated track.

2. Create narration

Write or paste a script, choose a voice and generate speech. Chinese uses dedicated Piper/VITS voices; English uses Kokoro; several European and Brazilian Portuguese voices use browser Piper models. You can also record narration with the browser microphone after granting permission.

3. Generate captions

Automatic captions use Whisper small q8 in a browser WASM worker. Timeline Studio adjusts coarse recognition timestamps toward nearby waveform energy. Caption text and segments are edited in the caption context panel while placement and styling stay global.

4. Build the timeline

Arrange visual, sticker, caption, source-audio, voiceover and music clips. Visual clips support transform keyframes, shape masks, speed controls, effects and transitions. Linked video and source audio can remain synchronized or be edited independently.

5. Preview and export

Preview uses the same project state as export. Choose resolution, frame rate, codec and quality. Timeline Studio records the rendered browser canvas, mixes active audio tracks and saves MP4 when supported, with WebM fallback.