From Text to Image: Kling 3.0 Omni Text-to-Video Storyboard Example

This entry demonstrates the basic usage of Kling 3.0 Omni on the text-to-video path: using only a single English prompt, the model turns a textual description into a continuous shot sequence with narrative pacing. On the schema side, only prompt and a few non-media parameters are exposed (duration, resolution, aspect ratio, native audio, smart multi-shot), so this is a minimal, purely text-driven example that contains no reference images or subject elements. For parameter choices, prefer_multi_shots is enabled so the model breaks down the shots on its own; combined with a 10-second duration and a 16:9 aspect ratio, this makes it easy to observe scene coherence and motion plausibility across shots; resolution stays at 720p to shorten generation time. The prompt is written in the order of "scene—subject—action—camera language—lighting atmosphere," which is a relatively stable description structure for text-to-video. It is suitable as a starting point for developers integrating text-to-video capabilities: on this basis, you can add elements (2–4 reference images of the same subject, referenced in the prompt as @name) to experiment with character consistency, or enable audio to test native audio and lip sync. Model family information shows that the Kling 3.0 series integrates video, image, and audio generation into a unified multimodal architecture, with cross-shot element consistency as a promotional highlight; this entry covers only the text-to-video entry point already declared in the schema, and other capabilities are subject to the official documentation.