Aspect ratio chosen as a preference before each generation
The help centre lists aspect ratio, number of outputs, model and generation length as the video preferences selected before generating. Four settings, all chosen per generation, with nothing stated about an episode. As of 2026-09-22.
| Point | What the vendor's page states | How this file reads it |
|---|---|---|
| The sentence | Aspect ratio, output count, model and length chosen before generating | Documented, as per-generation preferences |
| What it names | Four settings, and the moment they are set | Exactly what an operator controls on each run |
| What it leaves out | Any object larger than a clip | Consistency across a season is not this page's subject |
Inclusion rule. Three points per reading: the sentence as the vendor published it, the thing that sentence names, and what it declines to settle. Nothing here is inferred from an interface or a demo. Order. From the published sentence, to what it names, to what it leaves open.
1Preferences per generation put the discipline on the operator
Four settings chosen each time means four chances per clip for a season to acquire a shot in the wrong shape, at the wrong length, from the wrong model. Across forty episodes that is a filing job, not a feature.
Naming the model as one of the four is the part worth pausing on. A series generated across two model versions can drift for reasons no continuity mechanism will catch.
2Support material describes the next click, and that is its virtue
This is documentation written to answer a user standing in front of the interface, so the list is accurate and its scope is one generation. It is stronger evidence than a marketing sentence and a smaller claim than one.
The register records the genre alongside the content. Read next to pages promising a whole season, the honest narrowness here is easy to undervalue.
3Where this sits in the column
| Tool | What the vendor states about the deliverable |
|---|---|
| AI Drama Generator | An episode count, an aspect ratio and a language exposed as inputs |
| AI Dramas | Clips stated to generate in vertical 9:16 by default, 4 to 15 seconds each |
| Anijam | Every episode stated to export vertical at 9:16 |
| DramaSo | A captioned 9:16 episode at 720P or 1080P given as the output |
| Genra | 1080p HD on the product page; 9:16 vertical at the platform's highest resolution on the blog |
| Google Flow | Aspect ratio, model and generation length listed as preferences before a clip |
| Luma Agents | Reframing an aspect ratio and extending a duration listed as operations |
| Novi AI | Under a minute to several minutes stated, with no frame anywhere on the page |
| Pexo | 16:9, 9:16 and 1:1 delivered natively, up to 4K, from 5 to 120 seconds |
| PopShort AI | Video length given as an estimate that rises with the subscription tier |
| SceneMixer | 16:9 by default with 9:16 available, outlines written with cliffhanger pacing |
| Topview | Vertical 9:16 named as the format usually suited to short-drama platforms |
| TrendShort | A 9:16 vertical cut with captions and AI voiceover named as the export target |
Inclusion rule. Only entries whose own pages say something about the frame, the resolution or the pacing appear below. An entry silent on all three is left out of this column. Order. Alphabetical by tool name.
4The wording itself
- Serial formatAspect ratio, number of outputs, model and generation length are listed as the video preferences selected before generatingexposed as a preference
- Serial formatExtending video duration and reframing aspect ratios are listed among the operations available on the boardchanged by an operation rather than set once
5Sources
Read from the help centre article at support.google.com, checked 2026-09-22. The whole column is on serial format and this vendor on every column is on Google Flow; where the line falls between documented, claimed and absent is under how read. This vendor's other readings: the cross-episode cast. Other readings of the same column: SceneMixer, DramaSo.