A hundred and forty-three named rows against five unnamed

Synthesia publishes a table of 143 named rows for its translation feature. Kling AI states that its 3.0 video model produces native audio in five languages without naming them. The difference in what a reader can do with each is larger than the difference in size. As of 2026-09-22.

Inclusion rule. Two vendors chosen because one publishes the longest named list here and the other the smallest count. Rows are the questions each figure can answer. Order. Fixed order, most consequential question first.

1Twenty-eight times the size, and infinitely more searchable

A producer asking whether Vietnamese is covered can answer it in one search of the table and cannot answer it at all from the count. That difference does not scale with the numbers: a count of fifty would be equally useless for the question.

So the two cells are not two points on one axis. One is a set and the other is a magnitude, and the register keeps the distinction visible in the table rather than sorting both into a single column of numbers.

2The smaller claim is the more specific one

Five is attached to one named model and described as audio generated while the shot is produced. The 143 belongs to a translation feature that turns finished material into other languages, which is a different operation.

Comparing them directly would be comparing a dubbing inventory with a generation capability. Both are recorded in the same column because both are about the language a finished video speaks, and the row prints which feature each belongs to.

3Where each vendor is silent

Neither says anything about the language of a generated script or breakdown. One has nine declared site codes and the other twenty, so on the site column the order reverses and the vendor with five voice languages leads.

Between them the pair shows that all three columns move independently. A reader who took either vendor's strongest cell as a summary of its language position would be wrong about the other two in each case.

4Sources

Counted from the Synthesia supported languages table and read from the Kling AI 3.0 model guide. 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: CapCut and Fliki, Synthesia and HeyGen. The full set of pairings is here; single cells are indexed at entries.

• Voice languages
  A table headed “All languages and dialects supported for Synthesia translation feature”, running 143 named rows from Afrikaans to Zulunamed list publishedSynthesia Docs, supported languages / recorded 2026-09-22
• Voice languages
  Arabic appears as a base entry plus 13 named regional variants, from Bahrain to Yemen, inside the same 143-row tableone language's variants inside the listSynthesia Docs, supported languages / recorded 2026-09-22
• 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