ElevenCreative’s Chimera turns a still character into a talking video
By Steven Van ·
Chimera pairs Eleven v4 voices with Seedance 2.5 video, so an uploaded character image and a script become one emotional, lip-synced performance.
ElevenLabs launched Chimera inside ElevenCreative on October 21. It pairs two models, Eleven v4 for the voice and Seedance 2.5 for the video, so a character you create or upload can deliver its own lines with realistic, emotional speech.
How it works
Start with a character image or one of ElevenCreative’s avatars. Then either upload your own audio or write a script and pick a voice, or design a new one to match the character. You direct the performance with Eleven v4 audio tags such as [whispers], [laughs] and [yells], which control the timing and mood of each word.
Next, write a video prompt the way you normally would, describing the character’s movement, the setting, the lighting and the camera. Choose an aspect ratio and resolution, and Chimera generates the clip in one flow. In ElevenLabs’ words, Eleven v4 puts the emotion in the voice and Seedance 2.5 carries it onto the face.
Why it matters
Talking-character videos usually mean generating a voice in one tool, a video in another, and then fixing the lip sync. Chimera keeps the script, the voice direction and the video prompt in one place, so you can make dialogue scenes, explainers or ads with a consistent character from a single still image.
The announcement does not list Chimera’s pricing or credit costs.
Sources: ElevenCreative’s announcement and demo, ElevenCreative’s Chimera how-to.