Runway: a multilingual option with nothing named
The page that defines voices is the natural place for a language list, and this one does not carry it. A multilingual model option exists; no language is named beside it, so the field is recorded as unstated rather than counted. As of 2026-09-22.
| What the documentation settles | What it leaves to a take |
|---|---|
| A custom voice can be built from a sample of ten seconds to five minutes, at most 10 MB | Which languages that voice will then be able to speak |
| A voice can instead be described in words, with at least twenty characters | Whether a described voice can be asked for in a particular language |
| A voice becomes ready with a preview and is afterwards used by its own id | Whether the preview reveals the language range before a shot is made |
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.
1The absence is unusual because the page is about voices
Several entries skip this column because the document was written about pictures. That excuse does not apply here. A reference devoted to constructing voices, with sample windows, byte limits and an asynchronous lifecycle, still never says what those voices can speak.
So the blank is about editorial priority rather than about scope. The page answers how a voice is made and by what handle it is called, and treats the question of which words it can pronounce as somebody else's.
2A described voice makes the gap sharper
Asking for a voice in words rather than from a recording is the one route in this register that needs no performer. It also means the language is whatever the description implies, which is a very soft place for a market decision to rest.
The preview is the practical compensation. A voice reaches a ready state with something to listen to, so a production can at least hear what it got before committing a season to it. That is a test rather than a specification, and a test is what a blank column leaves.
3The other entries that never say which language
Eight more entries reach this column with nothing in it. Most were written about pictures; this one and one model card were written about the sound and still left the question out.
- LTX Studio — not documented by the vendor.
- Luma Ray — not documented by the vendor.
- MiniMax — not documented by the vendor.
- Sora 2 — not documented by the vendor.
- Veo — not documented by the vendor.
- Vidu — not documented by the vendor.
- Wan 3.0 — not documented by the vendor.
- Wan2.2-S2V — not documented by the vendor.
- Voice sourceA custom voice can be built from an audio sample between 10 seconds and 5 minutes long and at most 10 MBa published sample window
- Voice sourceA voice can instead be asked for in words, with the description required to run at least 20 charactersa route that needs no recording
- Per-character bindingA voice is created asynchronously, reaches a ready state with a preview, and is then referred to by its own ida stored voice object
4Sources
Read from the custom voices reference at docs.dev.runwayml.com on 2026-09-22. The same column across every entry is on languages; everything this vendor publishes about speech is on Runway. What counts as documented is on how read.