Logo
Image
Video
Voice
Apps
Motion Control
Avatar StudioNew
Marketing StudioNew
More
New
Other
Pricing

Turn your idea into video

Ready to create some AI magic? Describe your scene, pick your settings, and watch as AI brings your vision to life in stunning video.

Write a detailed prompt describing your video scene
Optionally upload reference images or videos
Hit generate and create cinematic videos in minutes
Start creating on the left
Alibaba

WAN 3.0 Edit — reference-driven AI video editing

WAN 3.0 Edit on Zyka is Alibaba's reference-driven video editing tier. Supply reference images, videos, or audio and describe the edit — WAN restyles, transforms, or recomposes with native audio.

  • Reference-to-video editing
  • Up to 30 second clips
  • 480p, 720p, and 1080p
  • Native audio generation
  • Adaptive aspect ratio

How WAN 3.0 Edit Works

01

Add your references

Upload reference images (up to 10), videos (up to 5, 15s total), and audio (up to 5, 15s total). Reference them in your prompt as @Image1, @Video1, etc.

02

Describe the edit

Write a prompt that tells WAN how to combine or transform the references — restyle a subject, swap a background, compose a new scene from the reference pool.

03

Pick output

Choose resolution (480p–1080p), duration (2–30s), aspect ratio, and whether to generate audio. Then render the edited clip.

About WAN 3.0 Edit

WAN 3.0 Edit is the reference-to-video edit tier of Alibaba's latest WAN 3.0 video model on Zyka. It takes a pool of reference media (images, videos, audio) and a text prompt and produces an edited video that respects character identity, visual style, movement, and sound cues across the references.

The edit workflow is well suited to character-driven edits, style transfer, re-composition and localized changes where you want the output to inherit specific attributes from the references rather than start from scratch.

For a faster render at a premium price, see WAN 3.0 Edit Prime — same input shape, same output quality tiers, speed-tuned endpoint.

Frequently Asked Questions

What is WAN 3.0 Edit?

WAN 3.0 Edit is the reference-driven edit tier of Alibaba's WAN 3.0 video model. Supply reference images/videos/audio and a prompt; it produces an edited video that follows the references.

How is this different from WAN 3.0?

WAN 3.0 is unified text-to-video, image-to-video, and reference-to-video. WAN 3.0 Edit is a dedicated reference-to-video tier focused on editing workflows.

What's the difference between Edit and Edit Prime?

Edit Prime uses a faster endpoint with a higher per-second cost. Edit uses the standard WAN 3.0 R2V endpoint — cheaper, slightly slower.

How much does WAN 3.0 Edit cost?

Billing is per second of output at fal rates + 20%: 6 credits/s at 480p, 12 credits/s at 720p, and 24 credits/s at 1080p. A 5s clip at 720p costs 60 credits.