Captions and formats

Podcli ships four caption styles and three aspect ratios. Captions scale to each canvas instead of floating mid-frame.

Every clip is captioned and cropped to a target aspect ratio. Pick these in the studio (New episode → Settings), with CLI flags, or via the create_clip MCP tool.

Caption styles#

StyleLook
brandedClean white text with the active word in a dark chip.
hormoziBold text with the active word punched in yellow.
karaokeWord-by-word fill as the line is spoken.
subtleUnderstated text for a calmer, editorial look.

Set it with --caption-style:

Formats#

FormatWhere it goes
Vertical 9:16TikTok, Reels, and Shorts. The default.
Horizontal 16:9YouTube and landscape players.
Square 1:1Feed posts.

Captions scale to each canvas, so text stays anchored instead of floating mid-frame. Set the format with --format:

Cropping#

Podcli frames the clip on whoever is talking.

  • center: a fixed center crop.
  • face: face-tracking crop with an exponential-smoothing camera.
  • speaker: mouth-motion speaker tracking for split-screen and Riverside-style layouts, with a scene-cut guard against jittery pans.