Text and image
Text and image to video
Both inputs, one clip. The image is the person. The text is the effect, and you do not have to write it.
Text and image to video is how the paid Seedance pass is built. Your photo is the identity. The template’s direction is the text. Seedance also receives a still of the sample so the move stays close to the card.
H3 takes the photo as the first frame and the same direction as text. Kling is the text-only model and is separate.