Scenario Generation API
Generate game-ready images, video, audio, and 3D assets from text prompts, reference images, or ControlNet guides. Supports text-to-image, image-to-image, ControlNet (txt2img and img2img), text-to-3D, image-to-3D, and modality-specific endpoints for video and audio. Accepts prompt, negative prompt, numSamples, guidance, numInferenceSteps, width, height, and modelId; returns an async job that resolves into one or more assets. A dryRun parameter returns the compute-unit cost without executing the job.