Video Generation

You can generate AI video asynchronously by instantiating the z-ai-web-dev-sdk to create a text-to-video task with your prompt, then using polling to check the task status until the video result is retrieved.

Yes, image-to-video conversion is supported. You can supply a starting frame and configure output parameters including duration, size, frame rate (fps), and optional audio for the generated video clip.

The best way to handle video generation task management in a backend service is using asynchronous polling provided by the z-ai-web-dev-sdk, which allows you to submit generation requests and retrieve results through status checks without blocking operations.

Yes, optional audio is supported during AI video generation. You can configure the output to include audio alongside your specified size, fps, and duration settings for marketing demos or educational visuals.

Asynchronous polling is used for AI video generation because rendering clips is time-consuming. It prevents backend service timeouts by separating task creation from result retrieval, ensuring robust task management for production environments.

You need to use the z-ai-web-dev-sdk for creating product demonstration videos when you require an SDK-first integration to programmatically generate marketing clips or educational visuals in backend applications with configurable outputs.