Hedra: full multi-language support, none of it named
Character 3 is described as text, image and audio to video with full multi-language support and a maximum duration of ten minutes. The duration is a figure a schedule can use. The language claim carries no list at all. As of 2026-09-22.
| What the documentation settles | What it leaves to a take |
|---|---|
| Described as text, image and audio to video with full multi-language support | Which languages that support covers, since none appears |
| A maximum duration of ten minutes is stated | Whether a face holds a viewer for ten unbroken minutes |
| Speech can be generated inline by naming a voice id from the voices endpoint | How many languages that endpoint's catalogue reaches |
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.
1A superlative adjective doing the work of a list
Full multi-language support is the strongest-sounding language claim in this register and the least usable one. It cannot be checked, cannot be planned against, and cannot be wrong, because no market is named and therefore none can be missing.
The more interesting reading is structural. This is an audio-driven product: the language of the output is the language of the recording that was handed over. Support for a language is, in that arrangement, mostly a statement about the lip movement rather than about a voice inventory.
2The voices endpoint is where the real answer would live
Speech can be synthesised inline by naming a voice id drawn from an endpoint, and an endpoint that returns voices is the natural place for a language to be attached to each one. The model page does not publish that, so a production that needs the list has to enumerate the endpoint instead of reading a page.
That is a reasonable engineering answer and a poor commissioning one. A decision about which markets a series can reach is usually taken by somebody who will never call an endpoint.
3The others that claim languages without naming any
Two more entries assert breadth without naming a market. One delegates the list to outside speech providers; the other names kinds of performance instead of places.
- LanguagesDescribed as text, image and audio to video with full multi-language support, and a maximum duration of ten minutesa claim carrying no list
- Voice sourceSpeech can be generated inline instead of uploaded, by naming a voice id drawn from the voices endpointa catalogue behind an endpoint
4Sources
Read from the Character 3 model page and avatar video guide at hedra.com on 2026-09-22. The same column across every entry is on languages; everything this vendor publishes about speech is on Hedra. What counts as documented is on how read.