What StoryShort calls a series, set against what a season needs
StoryShort's page defines a series as an automated feed of videos, one on Starter and four on Ultra, with per-video costs of about fifteen credits for a thirty-second video and about a hundred and ten for an AI movie. A season needs more than a feed, and the register's lines on hooks, episode length and persistence are the record. As of 2026-10-13.
1What this register already holds
Two short-drama makers' own lines from this register, beside what the pipeline documents about how an episode ends, how long it runs, and what crosses a join.
- Script to episodesThe AI is described as writing the script scene by scene, casting consistent characters and storyboarding every shot as hook, escalation, reveal and cliffhangera named sequence
- Assembling an episodeClips in which the characters speak their lines are described as assembled into one vertical episodestated as included
- Script to episodesA season of up to 60 episodes is described as plannable from a single premisewith a stated ceiling
- Serial formatThe stated output is a captioned 9:16 episode at 720P or 1080Pwith stated resolutions
- How an episode endsEvery episode is required to end on an unresolved beat, and the kind rotates: identity exposed, feeling suspended, understanding overturned
- Episode length and how it is builtVertical series usually run 60 to 120 seconds an episode, and a 90-second episode is roughly eight to twelve generated segments
- How many episodes a script yieldsScripts under 10,000 equivalent characters are read in one pass and longer ones in two passes up to 100,000; the episode count comes from the material
- Carrying state across a joinNew storyboards treat the join between two segments as a cut, and anything that must carry over, such as a bandage or a coat taken off, is written into the next segment
2What the page calls a series
StoryShort's pricing page sells automation: short videos with voiceovers, background music and auto-posting. Its word for a run is a series, and the page counts them — one series on Starter and four on Ultra — with the work described as an automated feed of videos rather than a story told in order.
Costs are quoted per video: about 15 credits for a 30-second video with AI images, about 110 for an AI movie. The pair is useful for a season because it prices the two ends of the format without promising an episode count at all, which is the number a serial production actually plans against.
3What a season needs that a feed does not have
The register keeps the working shape of a vertical season in its own entries: episodes usually run 60 to 120 seconds, and a 90-second episode lands at roughly eight to twelve generated segments. Every episode is required to end on an unresolved beat, and the kind of hook rotates — identity exposed, feeling suspended, understanding overturned.
None of that is a feed. A feed can run indefinitely; a season has to know how long an episode is, how each one closes, and what the next one inherits. The register keeps those as separate lines because a vendor can document the automation and publish nothing about the shape, and both positions are common.
4The second maker this page reads
DramaSo also writes for serials: a season of up to sixty episodes is described as plannable from a single premise, its stated output is a captioned 9:16 episode at 720P or 1080P, and its storyboard frames lock the look per line. Scope and shape are both published there, which is rarer than the automation this page is about.
Set the two side by side and the difference is what each page will commit to. One counts series and prices videos; the other names an episode ceiling and an output format. Neither is a schedule, and both are more than a feature list, which is what the register records.
5What the threads counted
The question the collection counted for this page is 19 near-identical threads asking what the best AI series is so far, filed under short drama. That is a question about finished work rather than published limits, and this register cannot answer it: it records what pages state, and a series that exists is evidence of a production rather than of a mechanism.
What it can do is narrow the question. A series needs three things at once — a cast that survives the run, episodes with an internal length, and an ending that pulls the next one — and the register's rows are the record of which of the three each tool's pages address.
6What each of them publishes about a whole run
| Maker | What its own page states about a series | Page read |
|---|---|---|
| DramaSo | Storyboard frames lock the look per line | 2026-10-13 |
| SceneMixer | “A short drama episode has two jobs: stop the scroll in its first seconds, and make the next episode feel necessary in its last.” | 2026-10-13 |
| StoryShort | An automated feed of videos (1 series on Starter, 4 on Ultra) | 2026-10-13 |
Inclusion rule. Rows are the makers this page is about, each read on its own public page, plus one pipeline whose season-scale steps this register records one cell at a time. A maker's cell carries the wording of the page read; the pipeline's cell is one sentence quoted from its published guides. A point a page does not address is written as not addressed rather than inferred. Order. Alphabetical by maker.
7What the threads asked
Under short drama the collection counted 19 near-identical threads asking which series stands up so far — a question about finished work rather than about any page's promises.
| Question | Threads with the same question | Theme |
|---|---|---|
| What is the best AI series so far? | 19 | short-drama |
Inclusion rule. Questions the collection counted, quoted the way they were written. Order. Descending by the thread count.
8When the other tool is the safer arrangement
- Automation is the point — Voiceovers, background music and auto-posting, with one series on the entry tier, answer a channel that needs volume rather than a story arc.
- The pricing has to be per video — About 15 credits for a 30-second piece and about 110 for an AI movie is a unit a season can be costed in without a plan-level conversion.
- The format is already settled — An automated feed suits a channel that has fixed its shape; a production still choosing episode length and hooks is better served by a page that states them.
Searches like "storyshort ai review" land here. Set beside it: cancelling MagicLight and Vidu past the one-minute mark.