Showcase: kinetic captions that hit on every word
Our name, then a short line about one call, video and no editor, one word at a time on black. Some land in yellow, one in our orange, the rest in white. It is the style you see all over short form video, where the words carry the energy as much as the voice does, and it is one of the first things people ask for when they hear we do captions.
How we made it
Each word is its own element on an HTML page with a start time. When its time comes it scales down from about twice its size and settles, which gives the "slam" feel. Color is a per word choice: emphasis words get the accent color, everything else stays white. Because it is HTML, the font is the real font, the kerning is real and nothing is baked into a template.
The timing is the important part. In a narrated video, every caption word is placed on the exact moment it is spoken, using the word timings from the voiceover. Here there is no voice, so the words are spaced evenly on the beat of the music bed.
How your agent can make one
Word timed captions are built into the explainer render. Turn on captions and voiceover and every spoken word gets its own timing:
POST /v1/explainers
{
"brandId": "<your brand kit>",
"aspect": "9:16",
"captions": true,
"voiceover": true,
"brief": "15 second vertical teaser. One line: one call, video, no editor needed. Big bold captions."
}If you already have a clip and only want it captioned in a style you like, point analyze_style at a reference video and then create_edit with your footage. That is the route for creators who want their own clips to look like a channel they admire.
What it costs
| Route | Credits |
|---|---|
| Explainer render with captions, under a minute | 12 |
| Or: analyze a reference style once | 2 |
| Then each edit in that style | 5 |
Voice is included in the explainer render. On the $10 starter pack, a captioned teaser is $1.20.
Get started and tell us which caption style you want to match.