Wan 3.0 AI Video Generator

Start with text or an image

Describe the subject, action, setting and camera movement for text-to-video. On a site that supports uploads, add a permitted image as the first frame for image-to-video. The keyframe workflow also supports an optional end frame when you want to guide the ending. XXXCut directs upload workflows to MeltVibe when uploads are unavailable on the main site.

Work with reference media

On upload-enabled sites, the Wan 3.0 Reference mode accepts images, videos and audio as references. Include at least one image or video; audio alone is insufficient. Reference mode is separate from first-and-last-frame generation. The composer shows the supported slots and limits, and reference video can increase the credit cost.

Choose settings and review credits

The Standard and Prime variants offer 480p, 720p and 1080p output with 5, 10 or 15 second durations. Other variants can have different settings. Check the options for your selected model in the generator and review the displayed credit cost before submitting. Availability comes from the current model catalog.

Keep your first prompt focused

Begin with one subject, one action and one camera direction. For example: an adult dancer turns slowly in a softly lit studio while the camera moves closer. Add a first frame when the starting composition matters, or use references when you need visual guidance. These are prompting suggestions rather than a promise of a particular result.

How to use it

1. Open the custom generator with Wan 3.0 selected.
2. Write your prompt. On an upload-enabled site, choose keyframes or Reference mode if you need image or media guidance.
3. Select the available duration, resolution and other settings, then review the credit cost.
4. Submit once, wait for the result and review the generated video.

FAQ