AI Video Editing Flows

AI recurring-character clip reels: measured cuts, length and loudness

AI recurring character IP (pets, oddballs, character universes): 8 pages and 60 reels measured, 6 reels torn down, 2 of them with a full teardown page. All figures were measured on 5 Nov 2026.

The format in numbers

The first group of rows comes from the Format scoreboard in our theme-pages workbook, which we recomputed from the raw files. The second group is measured on the torn-down reels of this format.

MeasureValueSource
Pages measured8Scoreboard
Reels measured60Scoreboard, all reels of these pages we collected
Median plays per reel37,912Scoreboard, median of the pages' median plays
Plays per follower, median across pages0.28Scoreboard
Reels at 3x or more of their own page median12 of 60Scoreboard
Posts per week, median across pages9.8Scoreboard
ConfidenceHigherScoreboard: set by the number of pages measured
Reels torn down6Our measurement of the files
Median length of torn-down reels15.3 sOur measurement of the files
Median hard cuts per minute5.6Our measurement of the files
Median big visual changes per minute6.8Our measurement of the files
Median integrated loudness-14.2 LUFSOver the 6 torn-down reels with an audio track
Median speech pace0.95 words per secondOver the 6 torn-down reels with speech

What we saw frame by frame

These counts come from the frame-by-frame notes of 4 reels of this format. A model read the contact sheets, and we counted keywords in its notes, so read the counts as approximate.

  • 3 of 4 reels: one continuous shot or no hard cuts.
  • 1 of 4 reels: text that stays still for the whole reel or most of it.
  • 3 of 4 reels: no text on screen or only a watermark or handle.
  • 2 of 4 reels: a viewer question or prompt in the post caption.
  • 1 of 4 reels: no end card.
  • 1 of 4 reels: sound effects or ASMR-style sound described.

Teardowns of this format

  • Teardown #13: a ten-second AI bodycam clip with one walk-up shot and a reveal in the last frames
  • Teardown #26: a ten-second AI kitten clip in one low-angle shot, with no text

Effects we have for this

Nothing in the notes of the teardowns in this format matches an effect on our list. Browse all our effects.

What we would test before copying it

  • Whether the reach holds on a page that is not already a hit. We found these pages through hashtags, which favour pages that already had one, and only 8 pages are measured.
  • Whether the editing pattern or the premise drives the plays. The data cannot separate the two.
  • Retention, shares and follows. These cannot be seen from outside Instagram.
  • Whether a page that posts at the median of 9.8 posts a week keeps its numbers over a longer run. We measured each page on its last ~10 reels only.

Method and limits

  • Plays, likes, comments, followers and posting dates were collected from Instagram by Apify on 5 Nov 2026.
  • For each of 69 pages we took the best reel and a typical reel of its last ten, 138 reels in all, and measured the files by script.
  • Ratio to the page median is plays divided by the median plays of the page's last ten reels, as measured on 5 Nov 2026.
  • Format statistics are recomputed here from the raw files and compared with the workbook's Format scoreboard; any difference is recorded under recheck.
  • Pattern counts per format are keyword searches over the frame-by-frame notes of that format, so read them as approximate.
  • Hard-cut counts miss dissolves, so a reel built on dissolves can show close to zero hard cuts.
  • The big visual change count also counts camera travel and scene changes, so it is not the same as the cut count. We show both.
  • Some full-HD files had no audio track, so loudness and speech are blank for those reels.
  • Plays include replays. Shares and follower conversion cannot be seen from outside.
  • Ten reels from one page are not ten independent tests.
  • Hashtag discovery favours pages that already had a hit, so breakout pages are over-represented.
  • Plays per follower favours small, newer pages; it measures reach health, not size.
  • Formats measured on only one to three pages are hypotheses.
  • No number here is a promise about what a new reel will do.

Plays and followers as shown on Instagram on 5 Nov 2026. The scoreboard rows matched our recomputation from the raw files on every field; the check re-runs the same arithmetic on the same files, so it does not test the method.

Page checked on Nov 4, 2026.