Video Gen is an eight-field framework for AI video generation: Scene describes what to depict, Motion specifies how things move, Camera controls camera movement, Lighting sets illumination, Style defines the aesthetic, Duration sets clip length, Negative lists exclusions, and Parameters adds model-specific flags. Works with Sora, Runway Gen-3, Pika, Kling, and Wan.
Video generation prompts add two critical dimensions over image prompts: Motion and Camera. Motion describes how subjects move within the frame. Camera describes how the camera itself moves. These two fields transform a static image description into a dynamic sequence. Getting them right is the difference between a compelling clip and a video that looks like a jittery photo.
Structured template for Sora, Runway, Pika, and Kling.
Use Video Gen for any AI video generation task. For static images, use the Image Gen framework.
Video Gen vs Image Gen: Image Gen has 7 fields for static images. Video Gen adds Motion and Camera for video output. Always use Video Gen for video generation — the Motion and Camera fields are essential.
Save your first Video Gen prompt in the structured editor and get a permanent REST API endpoint. Personal plan is free forever, no credit card required.
Start freeVideo Gen works with Sora (OpenAI), Runway Gen-3 Alpha, Pika 2.0, Kling 1.6, Wan 2.1, and Luma Dream Machine. The Parameters field handles model-specific settings.
Motion describes movement happening within the frame — a person walking, water flowing. Camera describes the movement of the camera itself — pushing in, panning, orbiting. Both are needed for compelling video prompts.
Most models produce best quality at 4–6 seconds. Start with 5–6 seconds and extend once you have a working prompt.