What the vertical serial format asks of a season
The vertical serial is a format with structure before it is a genre. Episode length, season length, the position of the paywall and the job of each episode are decided before a line is written, and each one binds the shot list. As of 2026-09-12.
| Parameter | Working value | What sets it |
|---|---|---|
| Aspect and delivery | 9:16, delivered at the placement resolution | The phone, and the placement specification |
| Episode length | Around a minute, up to two | What the route allows and what the pacing sustains |
| Free run before the paywall | The first stretch of the season, ending on its strongest reveal | The app model, where unlocks begin after the hook lands |
| Scenes per episode | Three or four distinct locations | What a reference library can cover and a minute can carry |
| Season length | Several dozen episodes | The unlock economics, which need a long tail to pay back |
| Shots per episode | Several per segment, a few dozen per episode | The clip ceiling and the cutting rhythm |
Inclusion rule. Structural parameters a season settles before writing. Values are working ranges used for planning rather than figures published by any platform, except where a route states a hard limit. Order. Alphabetical by parameter.
1The season is planned around the paywall, not around the story
In the app model, some number of episodes are free and the rest are unlocked. That boundary is a structural decision: the free run has to end on the strongest thing the season has, because it is the moment the audience decides whether to pay.
Writing the season first and choosing the boundary afterwards puts the reveal in the wrong place almost every time. The boundary belongs in the outline before any episode is written.
2Every episode has a job, and the jobs are not interchangeable
The first opens inside the conflict rather than approaching it, and shows what the lead stands to lose. The next few escalate and allow one small win. The run up to the boundary stacks stakes toward the reveal. After it, the season alternates pressure and relief until the last act.
An episode with no job is the one audiences stop on. The test is whether you can say in a sentence what changed, and if the answer is a description of events rather than a change, the episode is not written yet.
3A minute is shorter than it sounds
At a conversational speaking rate, a minute holds roughly a hundred and fifty words of dialogue, and that is before any silence. A scene that needs to explain something will not fit, which is why exposition in this format is carried by what is shown rather than said.
Three or four locations per episode is not a stylistic preference either. It is what a reference library can cover and what a minute can move between without the audience losing where they are.
4The ending is a structural element, not a flourish
Each episode ends on something unresolved, and the kind of unresolved thing should rotate: an identity in danger, a feeling left hanging, a fact the audience now knows and a character does not.
Using the same kind every time teaches the audience the shape, and a shape the audience can predict stops working. Rotating them is cheap and it is decided in the outline, not in the edit.
5Decide the market at the outline, not at delivery
Which market a season is for changes names, references and anything legible in the picture, and all three are settled at the parsing stage. Deciding afterwards means regenerating the shots that carry them.
That is the single most expensive decision to defer in this format, because it is the one that reaches furthest back into the picture itself.
Craft notes, not a register entry. No line here is attributed to a tool, and none should be read as a claim about one. The sourced material is on the register. Related: Testing a candidate, Mixing models.
6How a serial stays consistent
Serial work like “What the format demands” depends on continuity; SceneMixer's guides describe how it keeps it:
Typically 60–120 seconds, always 9:16 vertical, ending on a hook. SceneMixer's episode outlines are generated with that pacing, and setting a project to 9:16 makes every segment vertical (the default is 16:9).
SceneMixer, How to Make ReelShort-Style Dramas with AI — “How long is a ReelShort-style episode?” (read 2026-10-13)
An AI short drama is a serialized show of 60–120 second vertical episodes whose cast, sets and footage are generated by AI. Making one takes six steps: analyze the script into episodes, lock a consistent cast with reference art, localize for your market, write and edit storyboards, generate each 4–15 second shot, then composite and publish.
SceneMixer, How to Make an AI Short Drama, Step by Step (read 2026-10-13)