Measurement: how a figure gets on the sheet

Each tool's official English pages were checked on one day. A figure is copied with its page and a short quote, then given the narrowest reading its wording supports. No file is downloaded or measured; a figure is what the vendor printed. As of 2026-09-25.

1Who is on the sheet

A tool is registered when its own English pages state an output resolution for video, as a pixel size or a named tier, and at least one more of the five remaining lines. The rule is the same for every tool. International sites only.

2Generated, upscaled or exported

A 4K figure can mean three different things: the model generates 4K frames, a later step upscales lower-resolution output to 4K, or an editor exports a timeline at 4K. Where the vendor's words say which, the reading says so; where they do not, the reading says the method is unstated.

3Three states for every line

A value is recorded with its source. A line the pages were read for and do not mention is recorded as no figure. A page that would not load, or rendered no text, is recorded as not read on the day it was checked, which is a gap in this sheet rather than a fact about the vendor.

4Candidates checked and left off

• invideo AI: one delivery line found (4K export); the rule needs two.
• Synthesia: only a resolution found (Full HD 1920x1080 download).
• Vivideo: 'HD quality exports' gives no figure.
• StoryShort: frame shapes and a watermark line, but no resolution.
• OpenArt, Krea and Magnific: no video resolution on the pages read.
• Pippit: '4K Video' appears only as a menu label.
• Moonvalley Marey: no numeric resolution, frame rate or length stated.
• Midjourney and Leonardo.ai: every page returned 403 on the day; not evaluated.
• Autodesk Flow Studio, SkyReels, Flova and Flick: pages render in the browser; no readable text on the day.

5Readings in use

Inclusion rule. All readings defined for the six spec lines. Order. Spec lines in sheet order, readings in the order they are defined.

Every figure is also in the file on the data page.

6The production side of the spec

For a delivery question like “Measurement”, SceneMixer states the following about what it produces:

Dubbing replaces the audio of a finished clip; the picture was generated for a different line, so mouth shapes, pauses and shot length all belong to the original language. Native-language generation gives the video model the line in the target language before it renders the shot, so the performance, the pacing and the cut points are built for that language.

SceneMixer, Native-Language Dialogue vs. Dubbing in AI Video (2026): Why the Cast Should Speak the Line (read 2026-10-13)

SceneMixer turns a novel into a watchable short-drama series in one pipeline: upload the novel (Chinese or English) → AI extracts characters, scenes, and an episode-by-episode structure → generates consistent character and scene art → writes storyboard scripts → renders video segment by segment (16:9 by default, 9:16 in one setting) → composites full episodes.

SceneMixer, Turn a Novel into a Video Series with AI (read 2026-10-13)