Platform
Character (persistent)
A persistent character is an AI cast member defined once — face references, persona, voice — and reused across images, video and dialogue so they stay the same person in every scene.
Consistency is the hard problem of generated media; a persistent character solves it at the data level rather than by re-prompting.
On eroq
Characters at /studio/characters: reference photos, persona, a roster or cloned voice; @mention them anywhere; cast them per scene in Cinema; a Content kit of social photos; `char:<id>` in the API.
Questions
Can a character speak?
Yes — with its assigned voice in the voice studio, films and the API.
How many characters can be in one scene?
Up to four per render, brought in by @Name or the Add character button above the composer. In Cinema each scene carries its own cast with nothing inherited from the previous one, so a two-hander scene is not polluted by the crowd shot before it.
Does a character hold their face in video?
On the engines that read reference pictures, yes: the character's look sheet rides as a numbered reference picture — up to six on Seedance, four on Kling, three on Veo 3.1, Veo 3.1 Fast, Hailuo 3 and Wan 3.0 — and @Name in the prompt binds to it. On Motion One their picture becomes the start frame; text-only engines get the character described in words, with nothing to match.