Reference Image or Video to Video

Key Features of Reference-to-Video

Generate consistent character videos using reference images and videos as input. Maintain character identity and style across your video content.

Character Consistency

Keep the same character appearance and style throughout the generated video using reference inputs.

Multi-Reference Support

Use multiple reference images and videos to guide the generation with precise character control.

Flexible Output

Generate 720p or 1080p videos in various aspect ratios with 5 or 10 second duration.

How to Use Reference-to-Video

Follow these steps to create your first AI video

1

Upload References

Upload reference images and/or videos of your character or subject. Both images and videos help maintain consistency.

2

Write Your Prompt

Describe the scene and actions. Use @Video1, @Video2 to reference your uploaded subjects.

3

Configure Settings

Select resolution, aspect ratio, and duration for your video.

4

Generate & Download

Click generate and download your character-consistent video.

Pro Tips

Use close-up head shots with multiple angles for portraits
Upload both images and videos for best consistency
Use @Video1, @Video2 in prompt to reference specific characters

Writing Effective Reference-to-Video Prompts

Learn how to craft prompts that maximize character consistency.

Prompt Formula

@Video1 + Action + Scene + Style

Character Reference

Use @Video1, @Video2 to reference subjects from your uploaded videos

Action

Describe what the character is doing - movements, expressions, gestures

Scene

Set the environment, background, and lighting conditions

Style

Define visual style, camera angles, and cinematic techniques

Example Prompts

Good Example Example

@Video1 is walking through a busy city street, smiling at the camera. The sun sets behind skyscrapers, casting warm golden light. Cinematic tracking shot.

Multi-Character Example

@Video1 and @Video2 are dancing together in a colorful studio. They move in sync with upbeat music. Wide angle shot with dynamic lighting.

What Users Are Saying About Reference-to-Video

Sarah Chen

Sarah Chen

Content Creator

Wan AI has revolutionized my content creation workflow. The quality of generated videos is incredible, and the interface is so intuitive.

Mike Johnson

Mike Johnson

Marketing Director

Our marketing team now produces video content 10x faster. The AI understands our brand voice perfectly.

Lisa Wang

Lisa Wang

Educator

Creating engaging educational videos has never been easier. My students love the dynamic visualizations.

Frequently Asked Questions About Reference-to-Video

What is the latest version of WAN AI?

WAN 2.5 Preview is the latest version, featuring advanced audio capabilities that allow adding synchronized sound to videos. Compared to Google Veo 3, WAN 2.5 is more affordable, supports more formats, and delivers longer video clips.

What video quality can I expect?

Our AI generates videos in multiple resolutions up to 1080p, with options for 480p, 720p, and 1080p depending on your selected model and plan.

How long are the generated videos?

Generated videos are typically 5-10 seconds long, capturing the key moments described in your prompt. Some advanced models support longer durations.

How do credits work?

Credits are consumed based on the model type, resolution, and video length. Higher quality models and settings use more credits. You can purchase additional credits anytime.

Can I use generated videos commercially?

Yes, you own full commercial rights to videos generated with your credits. Check our terms of service for complete usage guidelines.