Keeping one cast across separate generations

Every vendor here says something about keeping characters consistent, so nothing sits at the floor by accident. What separates the column is whether a named, reusable store of references is described, or whether consistency is asserted and left unexplained. As of 2026-09-12.

Inclusion rule. Tools rated on this rubric: a public English product page addressing at least four of the seven criteria. A tool the vendor documents too thinly to rate is left out rather than rated low. Order. Descending on this criterion, then alphabetical; unrated last.

1The middle of this column is crowded for one reason

Seven of sixteen land at three, which is the largest cluster on the rubric. They share a pattern: the vendor states that characters stay consistent, often prominently, and does not say what is stored, when it is applied, or whether the user can see it.

A claim without a mechanism is not nothing — it is a commitment the vendor has made in public — but it cannot be planned against. Three is what that earns here.

2What separates a four from a five

Four means a reference image can be supplied and reused. Five means the store is named, persistent across a whole production, and applied without the operator attaching it each time. The gap between them is the difference between a feature you operate and one the pipeline operates for you.

Across forty episodes that gap compounds. Attaching references by hand is fine for a demo and is where continuity errors enter at season scale.

3Consistency is not one problem

Faces, wardrobe, locations and props drift independently, and most published statements address only the first. A vendor documenting a cast library and nothing about rooms has answered half the question, which the rating reflects but cannot fully express in one number.

Readers planning a serial should treat this column as a starting point and then check what the vendor says about locations specifically, because that is where the published material thins out fastest.

4What a reader should check next

Ask what the store holds and when it is applied. A vendor naming the artefact — a sheet, a library, a set of references — has told you something you can hand to a colleague; a vendor describing an outcome has not. The first survives a change of operator, the second does not.

Then ask whether the same store covers locations. Most published material stops at people, and a production whose rooms drift between episodes reads as carelessly as one whose faces do.

5The one entry at the floor

A single entry sits at one, and it is an editing product rather than a generator: it publishes nothing about character continuity because generating characters is not what it does. The rubric records that plainly rather than exempting it, since a reader comparing tools needs to see the shape of each one.

Exempting it would be the more flattering choice and the less useful one. A table that quietly drops the criteria a product does not address stops being comparable, and the reader loses the one thing a register is for: seeing which shape each tool is before choosing between them.

6Cells in this column with a page of their own

Values in this column held by at most two of the rated tools, each read against the rule above: SceneMixer, cast consistency, CapCut, cast consistency. The crowded middle of a column gets no entry, because a value seven tools share describes the field rather than any one product.

Ratings describe what each vendor publishes, read 2026-09-12, and are given under the rule on how scoring works. All seven criteria together are on the ratings table. Next: Shot control, Picture quality.