Status: Soon. Public video creation access is pending. The initial V1 path creates one five-second, 720p, 24 fps text-to-video clip using Wan 2.2. Combined multi-shot generation remains Soon.
Describe one shot
Write subject, action, setting and camera movement as one continuous moment. Example: A runner crosses a rainy pedestrian bridge at dawn; camera tracks beside them at walking speed; cool natural light; no cuts or captions. A five-second clip needs one clear action, not a list of scenes.
Review settings and submit
- When access is enabled, open /studio/video or choose Create → Video in the editor. Enter the brief under Describe your video.
- Inspect Model & output and choose from the options shown in your account. The initial V1 path is five seconds at 720p and 24 fps; longer durations or other resolutions are separate capabilities.
- Read the quote and balance. Press the generation button once and let the job finish. If submission status is uncertain, use Check job status or Retry original submission; do not submit a second paid request blindly.
- Review the playable result. Choose Download, Edit creation or Add to this editor as appropriate. Trim and place it in the manual timeline.
Plan several shots without implying a one-click film
Open Help me plan → Plan type → Storyboard draft and choose 2, 3 or 4 shots. Suggest a plan returns editable scene descriptions. Choose Use this scene for one clip, generate it, then repeat for another scene. Add the separate clips to the timeline and arrange them. The UI itself says combined multi-shot generation is still being completed.
Can I turn a still image into a video here?
This qualified guide covers text-to-video only. Reference or image-guided creation is Soon; do not assume an image input is accepted by the V1 path.
What if the clip is not what I intended?
Change one concrete detail in the prompt, check the next quote, and generate another clip. Use manual trim and sequencing for timing changes that do not require another model run.