PixVerse: a built-in voice, or one from your sample
Text to speech here supports built-in voices and custom voices created from supplied sample audio, and whichever is chosen travels as a speaker id on the generation request. Two routes, one handle. As of 2026-09-22.
| What the documentation settles | What it leaves to a take |
|---|---|
| Text to speech supports built-in voices and custom voices from user-provided sample audio | How long a sample the custom route needs, since no window is published |
| The chosen speaker id is passed as the lip-sync speaker on the generation request | Whether a custom speaker id persists between projects |
| Audio and video are each capped at sixty seconds and one hundred megabytes | Whether the sample is bound by the same ceiling |
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.
1One handle for two very different origins
A speaker id that might point at a stock voice or at a clone of a real performer is convenient in code and risky in a production. The request looks identical either way, so nothing in a log distinguishes a licensed catalogue voice from one built out of somebody's recording.
That matters for the same reason the distinction matters legally. A stock voice carries the platform's terms; a custom one carries whatever permission the sample came with, and only a production's own records connect an id to that.
2The published ceilings are for the call, not for the sample
Sixty seconds and a hundred megabytes are documented for audio and video on the generation request. Nothing states how long a sample the custom-voice route wants, which is the figure a casting decision needs, and it is the one figure this otherwise specific page leaves out.
Comparing across this column, published sample windows range from a single recording to twenty minutes of studio time. Without a figure here, a production cannot tell which end of that range it is being asked for.
3The others that hand a production a catalogue
Four more entries answer with something to choose from. This one and two others also accept a recording, so the catalogue is a starting point rather than the whole answer.
- D-ID — a voice id, or a recording by url.
- Hedra — a track, or a voice id from the endpoint.
- SceneMixer — a preset library, or the user's own recordings.
- Synthesia — a catalogue voice, or a cloned one.
- Voice sourceText to speech supports both built-in voices and custom voices created from user-provided sample audiocatalogue or supplied sample
- Per-character bindingThe chosen speaker id is passed as the lip-sync speaker on the generation requesta speaker id per call
- Length per callAudio and video are each capped at sixty seconds and one hundred megabytes
4Sources
Read from the speech and lip sync guide at docs.platform.pixverse.ai on 2026-09-22. The same column across every entry is on voice source; everything this vendor publishes about speech is on PixVerse. What counts as documented is on how read.