Actions3
Video → Delayed Inference
AI-generatedSummary
Initiate deferred video generation of an AI avatar speaking custom text, returning metadata with a URL for on-demand processing upon access.
Inputs
- Text (required) — The text for the avatar to speak.
- Avatar (required) — Select the avatar character to use for the video.
- Emotion (required) — Choose the emotion or expression of the avatar (Happy, Neutral, Serious).
- Language — Language code for speech synthesis (e.g., en, es, fr, de, ja).
Output shape
a single JSON object containing metadata including the URL where the video processing is deferred until accessed
The returned JSON includes video metadata but does not contain the video binary itself. The video generation is deferred and processed only when the provided URL is accessed.
Examples
Example 1: Create a delayed AI avatar video to be generated only on access
Set Text to spoken sentence, choose Avatar and Emotion, optionally set Language, then run.