Model
Text description of the desired video (or the motion/changes for image-to-video). Audio is generated too.
Input Image (Image-to-Video)
Optional input image URL. If provided, animates the image (image-to-video); otherwise text-to-video.
Reference Images (Reference-to-Video)
1-7 reference image URLs for reference-to-video (style/content). Tag them in the prompt as <IMAGE_0>, <IMAGE_1>, etc. Takes priority over Input Image. 480p/720p only.
Video length in seconds (1-15). Cost scales with duration.
Output resolution. Higher resolution costs more per second.
Aspect ratio (text-to-video only). Ignored for image-to-video (uses the input image ratio).