Choose the model
Open the generator with Wan 3.0 selected and confirm the workflow mode.
Create native videos up to 30 seconds long with richer reference inputs and more control over motion and visual consistency. Up to 30 seconds: Tell a fuller story in one clip.
AI VIDEO WORKFLOW
Create native videos up to 30 seconds long with richer reference inputs and more control over motion and visual consistency. Wan 3.0 brings text, image, video, and audio references into one video workflow and supports clips up to 30 seconds. Use those inputs to direct a complete scene, then refine or extend the result with the model's editing tools.
Explore the inputs, controls, and creative workflows available with Wan 3.0 on Sora 2.
Tell a fuller story in one clip. Plan the opening, main action, and ending before generating a longer clip. A clear sequence gives the model room to develop the scene without losing the subject or the narrative point.
Generate dialogue, music, and effects. Describe dialogue, ambience, music, or sound effects alongside the visual action. Native audio helps the scene feel coherent when voices and movement need to land together.
Guide scenes with images, video, or audio. Combine the reference types that best explain the target: images for appearance, video for motion, and audio for sound. State which parts of each source should be retained so the result follows a consistent direction.
Choose 480p–1080p and adaptive ratios. Select a supported resolution and aspect ratio for the final placement, including portrait or wide delivery. Keep the main subject inside the safe area when one scene must serve several formats.
| Capability | Wan 3.0 | Seedance 2.0 | Sora 2 Pro | Veo 3.1 Fast |
|---|---|---|---|---|
| Max Resolution | 480p, 720p, 1080p | 720p, 1080p, 4K | 720p, 1024p, 1080p | 720p, 1080p, 4K |
| Input Types | Text, Image, Video, Audio | Text, Image, Video, Audio | Text, Image | Text, Image, Video |
| Aspect Ratios | 16:9, 4:3, 1:1, 3:4, 9:16, Adaptive | 16:9, 4:3, 1:1, 3:4, 9:16, 21:9, Adaptive | 9:16, 16:9 | 9:16, 16:9 |
| Max Duration | 30 sec | 15 sec | 20 sec | 8 sec |
| Function Mode | General, Reference, Frames, Editing, Extend | General, Frames, Reference, Editing, Extend | Not specified | General, Frames, Reference, Extend |
| Seed Control | Not listed | Supported | Not listed | Not listed |
| Web Search | Not listed | Supported | Not listed | Not listed |
Get started in three clear steps. No editing experience is required.
Open the generator with Wan 3.0 selected and confirm the workflow mode.
2Write a focused prompt and upload any source media required by the workflow.
Review the settings, generate, then refine the prompt or model for another variation.
See how prompts, references, formats, and model choices become practical creative directions.
Wan 3.0 helps solo creators and teams explore more directions without adding more production tools.
Develop social concepts, series formats, and visual experiments faster.
Create campaign directions, channel variations, and localized assets from one workflow.
Prototype product media, launch concepts, and seasonal variations before full production.
Use AI for previsualization, mood development, storyboards, and option exploration.
Create native videos up to 30 seconds long with richer reference inputs and more control over motion and visual consistency. Wan 3.0 brings text, image, video, and audio references into one video workflow and supports clips up to 30 seconds. Use those inputs to direct a complete scene, then refine or extend the result with the model's editing tools.
Open the generator, choose Wan 3.0, add a prompt and any required source media, review the settings, and generate. Plan the opening, main action, and ending before generating a longer clip. A clear sequence gives the model room to develop the scene without losing the subject or the narrative point.
The available inputs appear directly in the generator and change with the selected model and workflow mode. Combine the reference types that best explain the target: images for appearance, video for motion, and audio for sound. State which parts of each source should be retained so the result follows a consistent direction.
You can use the settings exposed by the model, including supported aspect ratios, resolution, duration, and reference limits.
Yes. Sora 2 keeps leading models in one workspace so you can compare inputs, formats, and workflow modes.
Rights can depend on the model, source media, plan, and local law. Review the terms and use only assets you have permission to upload.
Write a concrete brief with subject, composition, action, camera or lighting, style, and final use. For edits, state what must stay unchanged.
Compare other AI models available in the Sora 2 creative workspace.
Start creating with Wan 3.0 in the Sora 2 workspace.
Try Wan 3.0 NowCreate native videos up to 30 seconds long with richer reference inputs and more control over motion and visual consistency.