Wan 2.1 Image to Video

Wan 2.1 Image to Video creates a generated clip from one opening image and a motion prompt. The still supplies the scene and subject appearance, while the text explains what should happen after that first frame. It is a straightforward image-animation workflow within the Wan 2.1 family.

The route targets compact 480p or 720p outputs where the task is establishing movement from an existing visual. Guidance strength can tune how tightly generation follows the written direction. The result is delivered as a new video asset.

Highlights

• High visual quality from images. The exact route is presented as generating high-quality video from an image. The supplied frame anchors the sequence.
• Motion diversity. Wan 2.1 explicitly emphasizes varied motion in image-to-video results. A starting still can support different action concepts.
• Dedicated image-to-video evaluation result. In a dedicated image-to-video human evaluation, Wan2.1 led every compared model. This statement is limited to that published comparison.
• Natural-looking motion from a single still. Wan 2.1 produces fluid, natural-looking video and realistic motion from one input image, supporting scenes that should feel animated rather than simply displaced.

Pricing

When to Use

Prompt Guide

Describe the motion and camera behavior that should emerge from the supplied frame.

Subject action: [movement]
Camera: [shot and motion]
Environment: [moving elements]
Pacing: [tempo]
Ending: [final framing]

{
  "image": "https://example.com/opening-image.png",
  "prompt": "The cars race forward in slow motion as the camera pans beside them and settles into a wide view.",
  "resolution": "720p",
  "guide_scale": 5,
  "aspect_ratio": "16:9",
  "seed": 42
}

Technical Specs

Related

• Wan 2.1 First/Last Frame to Video — use two boundary frames when both ends are defined.