Build your request
Only the parameters that apply to video-to-clips are shown. Set values on the left, copy the payload on the right.
source.urlrequiredThe source content URL: audio file, video file, YouTube link, Spotify track, article page, or PDF depending on the workflow. http(s) is enforced; bare domains get https:// prepended. For character-swap, the video whose people are replaced (a file URL, or a YouTube, TikTok or Instagram link).
{
"workflow": "video-to-clips",
"source": {
"url": "https://www.youtube.com/watch?v=ip30oRP9BP0"
}
}Free, no API key needed.
Good to know
- →source.url is the link of one YouTube video (youtube.com/watch?v=... or youtu.be/...), not a channel or a playlist.
- →By default the AI picks the best moments: options.outputCount sets how many clips (1 to 3), options.clipDuration their target length, options.searchQuery what to look for.
- →options.clipRange clips one exact time range instead ("start:end" in seconds).
- →voice.enabled = true makes an AI summary instead of clips: an AI voice (voice.voiceId) summarises the video over extracts of it.
- →media.search = true adds B-roll to the clips: media.bRollType places it, media.bRollFrequency sets how often it appears, media.type what it is made of.
- →options.videoEditor = true has the AI video editor turn the clip's key lines into motion graphics with camera moves. It replaces B-roll and has its own flat price. options.videoEditorStyle picks its look (clean, bold or paper), options.videoEditorAiPictures adds AI pictures where the talk calls for one, options.videoEditorSoundEffects its short sounds.
- →A time range, the AI summary and the AI video editor each produce a single video.
- →Each clip is its own project: the response returns the first clip's pid, and the webhook is called once per finished clip.
- →captions.autoCrop defaults to true for this workflow, so clips are auto-reframed. Set it to false to keep the source framing.