Sentioscope

Speech and voice controls, as each vendor documents them

HeyGen: a count that belongs to translation

Thirty or more languages are documented, and the place they are documented matters: the figure sits on video translation, with voice cloning and lip-sync described alongside it. That is a statement about a workflow, not only about coverage. As of 2026-09-22.

Reaching thirty markets by translating, not by choosingThe language count sits on video translation rather than on the speech endpoint, which describes a workflow: one master, one performance, and a localisation stage at the end rather than a language chosen before anything is made.One masterA single episodegenerated andapproved in itsoriginal language.Then translateTranslationThirty or morelanguages, withcloning and lip-syncnamed together.Per marketDeliveryVersions that keepthe originalperformer's voiceacross markets.
Fig. 1 Concentrating the work in one master also concentrates the risk, because a flaw in it reaches every market at once.
HeyGen on languages, statement by statement. Read from the vendor's API quick start on 2026-09-22.
What the documentation settlesWhat it leaves to a take
Video translation is documented for thirty or more languages, with cloning and lip-syncWhich thirty, and how the cloned voice holds up in each
A clone is instant from one recording, or professional from twenty minutes or moreWhether the instant grade survives being translated
A text to speech endpoint turns a script into speech audioWhether that endpoint reaches the same language count

Inclusion rule. Only statements found on the vendor page named in the caption, and only ones bearing on this one field. Nothing here is filled in from generated output. Order. In the order the statements arrive on the vendor's page.

1Where a count is published tells you what it is for

Attaching the figure to translation says the languages are reached by translating something already finished, rather than by choosing a language before anything is made. Those are two different products and two different places in a schedule.

For a series the first arrangement is usually the cheaper one: one master, one performance, and a localisation stage at the end. It also concentrates the risk, because a flaw in the master reaches every market at once.

2Cloning and translation in the same sentence is the interesting part

Naming voice cloning next to the language count implies the translated version keeps the original performer's voice. That is the feature productions actually want from localisation, and it is the one that raises the sharpest consent question, because the voice being carried into thirty languages belongs to someone.

The page states the capability and stops there. Which languages, how faithfully, and under what permission are all outside it, and only the first of those three could ever be settled by publishing a list.

3The others that publish a total with no names

Two more entries answer this column with a number. The three totals are five, thirty or more, and ninety-five or more, and the largest of them belongs to the model with no voice inventory at all.

  • Kling AI — a count, no names.
  • sync-3 — a count, ninety-five or more.
  • Languages
    Video translation is documented for thirty or more languages, with voice cloning and lip-synccounted on the translation endpointHeyGen, API quick start / recorded 2026-09-22
  • Voice source
    A voice clone is instant from one recording or professional from twenty minutes or more, and is then passed as a voice idtwo enrolment routes with a thresholdHeyGen, API quick start / recorded 2026-09-22

4Sources

Read from the API quick start at developers.heygen.com on 2026-09-22. The same column across every entry is on languages; everything this vendor publishes about speech is on HeyGen. What counts as documented is on how read.