Travel stories with a beginning and ending
Take the viewer from a street-level detail to a wider location reveal. Keep one clear route through the scene so a longer clip develops naturally instead of repeating the same motion.
Wan 3.0 AI Video Generator
Build a scene from a written brief, animate an opening image, or guide a shot with reference material. Create clips up to 30 seconds with optional native audio.
- 2–30 second clips
- 480p, 720p, and 1080p
- Optional native audio
0 / 20000
Creative directions
Loading preview…
Built for your workflow
Product demonstrations in context
Show an object being picked up, used, and returned to the frame. Establish its appearance with an image, then describe the interaction and the surface or setting around it.
Atmosphere for a larger edit
Create a quiet establishing shot, environmental movement, or a transition between scenes. Plan a few steady moments at either end so the clip is easier to cut into a sequence.
Choose text, an opening frame, or references
Text-to-video is useful when you are still designing the scene. Name the subject, the action, the setting, and one camera move. For image-to-video, use a clear opening frame and describe what changes after that moment.
Reference mode accepts images, video, and audio through the controls provided here. Explain what each input contributes rather than asking the model to reproduce every detail. Pixwit does not offer document or web-page uploads in this workbench, even though Alibaba offers those workflows in its own products.
Wan 3.0 and the earlier Wan video workflow
Wan 3.0 gives a continuous scene more time to develop. Alibaba documents a maximum of 30 seconds compared with 15 seconds for the preceding Wan 2.7 video model. That extra room is useful for an entrance, an interaction, and a reveal within one shot.
Wan 2.7 Image is a still-image workflow, so it is not a like-for-like video comparison. If you start from a generated still, bring that image into image-to-video and evaluate motion separately from the image quality.
How to create a Wan 3.0 video
Start with a small, testable scene before expanding the length or reference set.
Set the scene and input mode
Use text for a new composition or upload an image when the opening frame is already defined. Keep the intended subject visible and leave room around it for the movement you want.
Describe events in order
Write an opening moment, one main action, and an ending. Add camera direction and lighting. If native audio is enabled, describe the ambience, dialogue, or effects that belong to the action.
Choose the delivery format
Select duration, resolution, and aspect ratio to suit the destination. Start with a short draft when movement is uncertain. Check the updated credit quote after changing settings or adding reference video.
Review continuity before refining detail
Watch for changes to the subject, object shape, and camera path. Fix the most visible problem first, then test again with the same brief before increasing the output length.
Example prompt
9:16 | A travel creator opens a paper map at an outdoor café, looks toward a mountain path, and walks out of frame. Slow camera push, natural daylight, café room tone and footsteps, no readable map labels.
Wan 3.0 FAQ
Build your next scene with Wan
Begin with one action and refine the take.