September 3, 2026 · 8 min read
Most advice about short-form format is either a list of numbers with no explanation or a rule that was true two years ago. This is the practical version: what each constraint actually is, why it exists, and which setting controls it.
Vertical 9:16 is the only ratio that fills the screen on Shorts, Reels and TikTok. Anything else is letterboxed, and letterboxing is the single clearest signal that a video was made somewhere else and reposted.
The exception is deliberate: a 1:1 square sometimes suits a feed post where the video sits inside other content. For a Short, always vertical.
In ClipPilot the aspect ratio is a setting on every video. Image videos support 9:16, 16:9, 1:1 and 4:5. Animated videos support 9:16, 16:9 and 1:1.
A 9:16 video at 1080 by 1920 is the standard. Uploading something smaller is not fatal, since the platforms re-encode everything anyway, but detail lost before upload is lost for good.
Animated videos have three quality tiers: Standard at 480p, HD at 720p and Full HD at 1080p. The pricing difference is real, at 25, 35 and 65 credits per second, so the honest guidance is that HD is the right default for almost everything. Standard is visibly softer on a phone held close, and Full HD is worth it mainly when the video will also be shown somewhere larger.
Every platform draws its own interface over your video, and it draws it in roughly the same places.
The workable rule is to keep anything that matters inside the middle third vertically, and away from the right sixth horizontally. If a viewer has to move the interface to read your text, the text may as well not be there.
This is also why the prompt guide recommends framing subjects high in a vertical shot. A subject centred for a landscape frame ends up partly behind the title bar once it is cropped to vertical.
YouTube counts anything up to 3 minutes as a Short. TikTok allows far longer. Reels sits in between. All three ceilings are much higher than what actually performs.
The useful number is not the limit, it is the drop-off. Retention falls steeply in the first three seconds and then again around fifteen. A video that has said what it came to say by second twenty finishes strong; the same content stretched to fifty seconds finishes weak, and finish rate is what the algorithms weight.
In practice: cut to the shortest length that still lands the idea. If you are unsure, it is too long.
Length is controlled differently by format. An image video follows the length of its script. An animated video is the scene count times the seconds per scene, which is also exactly what you are charged for, so there is no reason to pad one.
A large share of short-form is watched muted, at least for the first second, so captions are not an accessibility extra. They are how the video communicates before someone decides to turn the sound on.
Three things make captions work:
ClipPilot generates the voiceover and its per-word alignment in a single call and derives the captions from that alignment, which is the reason the highlight tracks the speech. There are two caption styles and a choice of highlight colour, both sized for a vertical frame.
On animated videos every clip is generated with its own sound. That audio is what carries the ASMR and found-footage formats, which is why several of the templates deliberately have background music turned off.
When there is a voiceover, the clips' own audio is ducked underneath it automatically rather than competing with it. When there is no voiceover, it plays at full volume.
Background music, where you use it, should sit low enough that the narration is never fighting it. The volume is adjustable per video.
Every constraint above matters less than this one. Something has to move, and something has to be interesting, before the third second.
For narration-led videos that means opening on the most specific thing you have, not on context for it. For visual formats it means the format itself has to be legible immediately, which is why the templates are all built around shapes that read in a single frame.
9:16 vertical. It is the only ratio that fills the screen on Shorts, Reels and TikTok. A 1:1 or 16:9 video will still upload, but it gets letterboxed with dead space above and below, which reads as reposted content rather than native content.
Up to 3 minutes. That is the ceiling, not a target. Retention on short-form falls sharply with length, and most Shorts that perform run well under a minute.
In the middle third of the frame, well clear of the bottom. The bottom fifth of a vertical video is covered by the title, the channel name and the action buttons on every platform, so captions placed there are partly hidden.
Yes, if it is 9:16 and has no burned-in platform branding. The safe zones differ slightly between apps, so keeping important content out of the bottom fifth and away from the right edge covers all three.
The glass-fruit format explained: why it works, how to pick a subject and a material that render cleanly, and how to keep a series looking consistent.
A walkthrough of the entire process, from writing a script to downloading your finished video.
Turn a script or a one-line idea into a finished vertical YouTube Short with AI visuals, voiceover and synced captions. No editing, no camera, no watermark.