Feature
Animated captions that follow every word
Captions carry the video for everyone watching on mute. Ours know when each word is said, so they can move with the voice.
Our AI transcribes the video and keeps the time of every word, so the captions can animate with the speech: the spoken word lights up, fills like karaoke, or pops in one word at a time. Pick a style, a font and colours once and every caption follows; retype any line and it keeps its timing.

Five styles in the editor
Every style works with the same word timings, so you can switch between them at any point without transcribing again.
| Style | What it does | Good for |
|---|---|---|
| Highlight | Colours the word being said | Talking heads, tutorials |
| Tag (box) | Puts the spoken word on a coloured tag | Bold short-form, product videos |
| Karaoke | Fills each word as it's said | Storytelling, faceless narration |
| One word | Shows one big word at a time, popping in | Fast hooks, the punchy short-form look |
| Clean | Plain text lines, no animation | Interviews, calm explainers |
The look, once for all of them
Choose the text colour and the spoken word's colour, the font (Grotesk, Impact, Rounded, Serif or a marker hand), the size, all caps or not, and a shadow or a box behind the line. Drag one caption on the video to move every caption.
Or say it in the chat: "yellow highlight, all caps, a bit bigger, higher up" restyles every caption at once.
- Captions set to one word at a time
- Text white, spoken word #FFD84D
- All caps on
Captions on script videos
When our AI makes a video from your script, the captions come from the narration's own word timings: karaoke and pop are outlined over a soft dark band so they read over any picture, and clean puts each line in a rounded pill.
Motion graphics and big words are kept in a box clear of the captions' band, so a chart or a title never sits under the subtitles.
Fix a line without losing the timing
Transcripts get names and jargon wrong. Retype a caption and its words keep their place in time. Add a caption by hand for something the transcript missed, and it goes on at the playhead and makes room for itself.
Translate the captions into another language from the chat, in place, when you want the same video for a second audience.
Where captions sit
On a vertical video the captions stay inside the safe area: clear of the top 220 pixels and the bottom 360, where TikTok, Reels and Shorts put their own buttons and captions. Change the format and they move into the new frame's safe area.
Questions
Are the captions word by word?
Yes. Every caption keeps the time of each word, which is what the karaoke, highlight, tag and one-word styles animate.
Can I make captions like the big viral ones?
The one-word style with all caps, a bold font such as Impact and a coloured spoken word gives the punchy short-form look. Tag puts each spoken word on a coloured box.
Do captions burn into the export?
Yes. They're drawn into the MP4 exactly as the preview shows them, so they look the same on every app.
Can I edit a caption's words?
Retype any line and it keeps its timing. You can also add a caption by hand at the playhead or ask the chat to fix a word everywhere.
Which languages work?
Captions follow the language that's spoken. Our AI can also translate the captions and on-screen text into another language in place.
Keep reading
- Chat editingType the edit in plain words: cut the pauses, add captions, swap the music, add a lower third. Our AI makes it on a real timeline, with one Undo.
- AI voiceoverNarrate any script in one of 130 AI voices, or your own recording. Speed from 0.8x to 1.25x, and captions that follow every word of the voice.
- TikTok videosDrop in your clips and say the edit: cut the pauses, add one-word captions, put a hook on the first second, keep it clear of TikTok's buttons. Then export.
- YouTube ShortsWrite a 15-second fact and get a finished vertical Short: 1080×1920, a picture for every line, a narrator and big captions. Then make the next one.
- Script to videoPaste a script and our AI makes the video: a voice, a picture and motion graphics for every sentence, word-timed captions, up to 15 minutes long.
- Silence removerCut every pause, um and retake from a talking video in one go. Our AI finds the dead air from the words and the sound, and you can still change every cut.
Updated 29 October 2026
Make your next video
Upload a video or paste a script, then say the edit. Our AI makes it on a timeline you can still change.
Try for free