Templates
A template is a preset of styling and clipping logic. It's applied to a video on upload and carries over to all its clips.
Create a template
Templates → New Template → name it → configure the blocks below → Save.
Styling
Background
What sits behind the clip's square frame on the 9:16 canvas:
| Mode | What it does |
|---|---|
| Full screen | clip cover-fills the whole canvas (centered) |
| Black bars | clip square centered, black above and below |
| Self-blur | a blurred copy of the video fills the area around the frame |
| Custom asset | a background video from your assets |
Banner
An image or video (PNG, GIF, APNG, WebP, MOV) overlaid on the clip — logo, CTA, subscribe prompt. Position is set in the preview.
Music
A background track from your assets. Volume is set when uploading a video.
Subtitles
5 animation styles:
- Karaoke — highlights the active word
- Word Pop — one big word at a time
- Sentence — the full phrase
- Spotlight — outlined active word with a smooth slide
- None — no subtitles
Plus subtitle color and font from a bundled set (Montserrat, Evolventa, TT Norms, HeadingNow, various weights). Each font in the dropdown is shown in its own typeface.
Show moment caption
A toggle. The moment caption is the short AI hook text at the top of the clip. Turn it off if you don't want it.
Preview Video
The Preview Video button opens the 9:16 canvas editor. Drag elements:
| Element | Control |
|---|---|
| Clip frame | body — move vertically; corners — resize (square) |
| Banner | move in any direction |
| Caption | move vertically |
| Subtitles | move vertically |
Resize the frame by dragging any corner — the clip gets bigger or smaller, with the chosen background filling the space around it. Save applies positions and size.
Segmentation (clipping logic)
Controls what and how the AI cuts:
| Field | What it sets |
|---|---|
| Number of clips | target fragment count (1–20). Empty = 5 |
| Min clip length | minimum duration in seconds (5–300). Empty = 30. Hard rule |
| Output language | language of captions and comments. "Auto" = match the video |
"Auto" detects language from the transcribed speech — an English podcast yields English hooks, a Russian stream yields Russian ones.
Custom LLM prompt (optional)
An advanced field that overrides the moment-selection instructions. Empty = default prompt. Format — see the clip_count, min_clip_duration_sec, output_language placeholders in the hint.
Default template
Set in Settings — applied to all new uploads automatically.