p-video-avatar
Pruna Video Avatar is the fastest talking-avatar / lip-sync model on the market — drop in a portrait photo plus either an audio clip or a written script, and get back a video of the subject speaking with synchronised lip movement, in 30 voices and 10 languages.
Input
Portrait photo of the avatar's first frame. JPG, JPEG, PNG or WebP.
Optional audio clip the avatar should lip-sync to. When provided, voice_script, voice and voice_prompt are ignored.
What the avatar should say. Required when no audio file is uploaded; the script is read aloud in the selected voice and language.
Voice timbre used when voice_script generates speech.
Language and accent the voice_script is read in.
Optional style hint for the voice (tone, pacing, accent, emotion). The text itself is not spoken — only the script is.
Optional visual direction for how the avatar moves and behaves while speaking.
Output video resolution. 1080p costs ~1.8× more per second than 720p.
Optional random seed for reproducible generations. Leave empty to randomise.
When true, prompts and the input image skip safety screening before generation. On by default for this model.
When true, the video_prompt is used verbatim without automatic enhancement.
Output
Run the model to see output here.
Pricing
Features
API Examples
curl -X POST \
/api/v1/model/prunaai/p-video-avatar \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"input": {}
}'