Wan-2.1 is an image-to-video model that generates high-quality videos with high visual quality and motion diversity from images.

Do not go beyond this recommended resolutions or your video will take forever to generate. Upload your 640x640 image for square or 832x480 in 16:9 (portrait or landscape). The default duration is set to 4s. WAN I2V is quite heavy (VRAM-consuming). It is recommended to load an image with fewer resolution/dimensions.

Write down your prompt. You can utilize GPT, Deepseek, or another prompt generator to achieve a scene that is desirable for WAN Image to Video.

The seed number represents the variation of the generated video.