Video Generation

Video generation from text prompts or images is handled by creating an asynchronous task and polling for completion to retrieve the final video URL. You provide text or image inputs, configure parameters like fps and duration, and the system returns the generated video.

Yes, AI video generation supports configurable quality, size, fps, duration, and optional audio. You can validate inputs and specify these parameters when initiating the asynchronous task to get a customized final video URL.

Asynchronous polling for AI video generation works by submitting a generation task and continuously checking its status until completion. Once the task finishes processing the text or image input, polling retrieves the final video URL for use in web apps or batch jobs.

Yes, the z-ai-web-dev-sdk is required as a backend dependency to generate AI videos. This SDK facilitates asynchronous task management, input validation for text or image URLs, and polling to return the completed video output.

AI video generation can be embedded into web apps, batch jobs, or prototyping pipelines for rapid media production. It handles multiple input formats like text and image URLs asynchronously, making it suitable for scalable video creation workflows.

Text-to-video generation creates videos from descriptive text prompts, while image-to-video generation uses input image URLs. Both methods support configurable size, fps, duration, and optional audio, returning a final video URL upon asynchronous task completion.