Twenty site codes and five voices, against fifteen and fifteen

Kling AI declares 20 site codes and publishes a voice count of five. SceneMixer declares 15 and names 15 voice languages plus Cantonese. Both also declare a Chinese alternate on another domain, and the rule discards both. As of 2026-09-22.

Inclusion rule. Two vendors chosen because their site columns are close and their voice columns are not. The last row records the identical rule effect on both. Order. Register columns in fixed order, then the discarded declaration.

1Five site codes apart, and a different kind of voice cell

On the site column these two are neighbours: twenty against fifteen, both reaching beyond Europe. On the voice column one publishes a set a reader can search and the other a number they cannot, which is a difference in kind rather than degree.

A reader who needs Ukrainian can settle it on one row in a second and cannot settle it on the other at all. Five unnamed languages and fifteen named ones are not four columns of the same table.

2The same rule takes a Chinese alternate from each

Both vendors declare a Chinese alternate pointing at a separate domain, and under the rule applied to everyone both are discarded. That is the clearest demonstration in this register that the rule is not tuned per vendor.

It costs one vendor a twenty-first code and the other a sixteenth, and both discards are printed on the vendors' own pages. A rule visible only where it helps somebody would not be worth writing down.

3The middle column separates them completely

One publishes nothing about the language of generated documents; the other states that output follows the interface setting and names the document types. That is the only filled cell in that column across the whole register.

Which makes this pairing the closest thing here to a like-for-like comparison of disclosure. Two vendors with similar site declarations, one of which has answered the other two questions and one of which has answered neither.

4Sources

Read from the Kling AI 3.0 model guide and the SceneMixer languages guide, with both hosts' markup counted. Both rows are kept in the register under one rule, set out on how counted; the column is described on dialogue language.

5Both of them against the rest of the column

Inclusion rule. Tools with a public English site selling AI video generation or AI video dubbing. Site languages count hreflang alternates served from the host being read; x-default and alternates pointing at another host are not counted, for every vendor alike. Order. Alphabetical by tool name.

Other pairings that turn on the same kind of difference: Vidnoz and Fliki, Runway and Higgsfield. The full set of pairings is here; single cells are indexed at entries.

• Voice languages
  Native audio on VIDEO 3.0 in five languages, with voices bound to elementscount published without a listKling AI, model guide / recorded 2026-09-12
• Site languages
  A twenty-first alternate is declared for Chinese and points at a different domain, so the counting rule skips italternate excluded by the rulekling.ai / recorded 2026-09-22
• Voice languages
  Spoken dialogue is set per project to any of the 15, with Cantonese a 16th option for dialogue onlyset per projectSceneMixer, languages guide / recorded 2026-09-12
• Site languages
  A sixteenth alternate is declared for Chinese and points at a separate domain, so the counting rule skips italternate excluded by the rulescenemixer.com / recorded 2026-09-22