Start with your source
Add a 2–15 second MP4 or MOV up to 64 MB, at 16 fps or higher, or choose a suitable video from your library.
Genjutsu AI. Your motion. A new reality.
Reimagine the character, the world, or the details.
Genjutsu AI
Your motion. A new reality. Start with your own video, then replace a character, background or object with a reference image, a prompt, or both.
Create Genjutsu AI videoHow it works
One source clip. A new visual direction.
Add your clip and describe the change. We prepare any reference images, then generate your video using the source performance.
Reimagine the scene
Add images for the characters, outfits or objects you want to replace. Use Prompt to identify each replacement and describe any background changes.
Automatic three-view preparation
After submission, each image becomes a front, side and back reference sheet on a clean single-color background. Your prompt guides the subject selection and appearance.
Generate your effect
Once all reference sheets are ready, your video is generated. Motion transfer follows the source performance; Objects swap focuses on the elements you select.
Optional visual reference
Use images to guide a character or product, and describe the background in Prompt. With only a source clip and description, generation skips image preparation.
Source-based output
Choose an available output resolution. The finished clip follows your source duration, and its aspect ratio is selected automatically from the source.
Credits before creation
The credit estimate includes reference-sheet preparation and video generation. The final quote uses the measured input-video length, output duration and resolution.
Review your result
Follow the generation status, preview the finished video and open the saved result. Reference-guided motion and visual details can vary.
What will you change?
Make the original performance your own
Keep your description focused on the change you want to see.
A new character look
Turn a dancer’s outfit into flowing ivory fabric, introduce a new character appearance, or explore a different visual style around the source performance.
A different background
Move the performance from a studio into a quiet sandstone courtyard, a neon city or another environment while guiding the same action and camera direction.
A fresh product scene
Identify the object to replace and add a product image if useful. Guide a new finish or appearance while keeping its interaction with the scene consistent.
Start free. Grow when you’re ready.
One credit balance for video, image, and voiceover creation.
Starter
For occasional creation
$24/month$29
Billed $288 yearly · Save $60
Models included
MiniMax H3
Seedance
FLUX 3Seedream
Nano Banana- 500 credits per month
- Up to 13 fast videos
- Up to 166 images
- UGC templates & Story creation
- Image generation & editing
- AI voiceovers
- Preview & download
- Credit top-ups available
Creator
For regular content creation
$65/month$79
Billed $780 yearly · Save $168
Models included
MiniMax H3SeedanceFLUX 3SeedreamNano Banana- 1,600 credits per month
- Up to 42 fast videos
- Up to 533 images
- UGC templates & Story creation
- Image generation & editing
- AI voiceovers
- Preview & download
- Credit top-ups available
Studio
For high-volume creation
$182/month$199
Billed $2,184 yearly · Save $204
Models included
MiniMax H3SeedanceFLUX 3SeedreamNano Banana- 5,400 credits per month
- Up to 142 fast videos
- Up to 1,800 images
- UGC templates & Story creation
- Image generation & editing
- AI voiceovers
- Preview & download
- Credit top-ups available
Credits are shared across images, videos and voiceovers. Monthly credits don’t roll over.
Frequently Asked Questions
What is Genjutsu AI?
Genjutsu AI is Clipin’s video reference effect for changing a character, background or object around the performance in your original clip. Choose Motion transfer to create from your clip’s performance, or Objects swap to edit selected elements.
What kind of source clip should I use?
Upload a 2–15 second MP4 or MOV up to 64 MB, at 16 fps or higher. Keep the subject visible and use clear motion. Both modes support 480p, 720p and 1080p output.
Do I need a reference image?
A source clip and description are enough. You can also add up to 10 JPG, PNG or WebP images, up to 10 MB each, for replacement characters, outfits or objects. Describe background changes in Prompt. Type @ to mention your images or video. If Prompt is off, add reference images to define the new look.
What happens to my reference images?
After submission, each image is used to create one three-view reference sheet: front, side and back on a clean, solid-color background. Your prompt helps identify the subject and requested appearance. Both modes use these prepared sheets to guide video generation, keeping the same @image numbers. If image preparation fails, video generation does not start. Unseen views are AI-generated interpretations.
Can I change more than one thing?
Yes. Describe the characters, background or objects you want to change, and use reference images to guide their new appearance. Mention which parts of the source should stay the same.
Will the motion stay exactly the same?
The source video guides the performance and camera movement. This is reference-conditioned generation, so motion, appearance and details can differ. Review the result before using it.
How are credits calculated?
The displayed credit limit includes three-view preparation for each reference image and video generation. Video pricing includes both input-video and output-video duration at your selected resolution. The final quote uses the measured clip length and never exceeds the displayed limit. With no reference images, there is no image-preparation charge. Completed image steps may still use credits if a later step fails.
Does it use Higgsfield’s Genjutsu model?
Both Motion transfer and Objects swap use Wan 3.0 through Clipin. Reference-image preparation uses Seedream 5 Lite. The sample image is an AI-generated concept created for Clipin.
What happens to the sound?
Both effects keep the original video’s soundtrack by default. In Motion transfer, enable Regenerate audio to use a new model-generated soundtrack instead. Objects swap preserves the original soundtrack. Motion transfer can optionally use the source audio as a reference for performance timing and lip movements. This is off by default. With Regenerate audio off, a silent source stays silent.
Create your Genjutsu AI video
Upload a source clip, describe the change and bring a new scene to life.
Create Genjutsu AI video