Make an explainer with an AI voiceover and AI visuals

An explainer is a script with pictures. When the voice is generated, the script decides the timing, so the order of work is voice first, then visuals, then captions and music.

Write the script, then generate the voice

Split the script into beats of one or two sentences, each carrying one idea. A beat is the unit you will illustrate, so a beat that holds two ideas needs two pictures. Read it aloud once, and shorten any sentence that is hard to say.

1. Open the AI Gen tile on the left panel, switch to Audio, then open Voice.
2. Choose a model and a voice. Press the play button next to the voice to hear a sample.
3. If the chosen model offers Voice direction, describe the delivery or open the presets menu. The presets include Warm, Documentary and News anchor.
4. Paste the script into Script and press Generate audio. The voiceover is added to the soundtrack at the playhead, and it lowers other audio while it plays.
5. Listen back. If a line needs work, edit the script and generate it again.

Generation runs on your DepCut account and spends credits, and several can run at once.

Make one visual per beat

1. Switch to Image. Write a prompt for the first beat and press Generate image.
2. For the next beats, type @ in the prompt to reference an earlier image, so the subject keeps the same look from shot to shot.
3. Switch to Video for a beat that needs motion. Choose the Duration and Aspect ratio and press Generate video. Renders take a minute or two, and you can keep editing meanwhile.
4. Drag finished image tiles onto the timeline. For a video tile, hover over it and click + (Add to timeline).

Line the pictures up with the voice

1. Open Subtitles and press Generate subtitles. The captions are timed to the voiceover.
2. Open the Content tab and click the timestamp on the first caption of a beat. The playhead jumps to that moment.
3. To start a picture at that moment, select the picture before it, open Trim and drag Trim end until the next picture begins at the playhead. An image has no source length to run out of.
4. Press the up and down arrow keys to jump between clip boundaries and check each hand-off.

Finish with captions, music and export

Pick a caption style in Styles. Minimal and Clean keep the attention on the pictures. In AI Gen, open Audio then Music, choose Instrumental, and add a track from the Sample library or press Generate music. Because the voiceover lowers other audio by default, the music stays underneath it. Fine-tune this with Duck on the voiceover clip.

Open Details for the Caption and Tags, then open Export and press Export video. The finished file appears under Exports in Media.

Let the Agent plan the whole thing

Press ⌘J to open the Agent and describe the video you want. It writes a script, breaks it into shots and shows you the shot list. It waits for your approval before it renders, because every shot spends credits. Review the list and change what you need before you approve.

Your turn

Write your first three beats, then open DepCut and generate the opening voice line.