Keeps the camera, replaces the subject
The source clip sets the shot: the pans, the pacing and the timing of each action stay in place. Only the person or product you name in the prompt is replaced with the subject from your images.
Kling 3.0 OmniNew
Video Swap
Swap the subject of a video with your own
Replace a person or product in an existing clip with the subject from your images, keeping the original camera movement and timing.
Try Video SwapSpecs
- Output
- Video
- Inputs
- Text · Image (up to 4) · Video
- Aspect ratios
- Auto · 16:9 · 9:16 · 1:1
- Resolutions
- 720p · 1080p
- Native audio
- Yes
- Price
- From 264 credits
FAQ
Questions, answered
What do I need to upload for an AI video swap?
One source video, 1 to 4 images of the new subject and a prompt. The prompt is required for Kling 3.0 Omni; it tells the model what to replace and what to keep.
Can I replace person in video and also a product?
Yes. The subject can be a person, a character or a product. Describe the target clearly in the prompt, especially when several people or objects appear in the frame.
How many credits does it cost, and what if a generation fails?
Pricing starts at 264 credits and depends on resolution and duration. The exact cost appears before you start. Failed generations are refunded automatically. New accounts get 50 free credits, and the Free plan adds 50 credits every month.
Can I use the result commercially?
Content you generate can be used in ads, stores and social media, subject to NeoMundo's terms of service and the model provider's policies. Only use footage and likenesses you have the rights and consent to use.
Which aspect ratio should I pick?
Auto is the simplest choice. Pick 9:16 for TikTok, Reels and Shorts, 16:9 for YouTube and web pages, and 1:1 for square feeds and product listings.
How do I get a cleaner swap?
Use a source clip with one main subject, even lighting and little blur, and subject images that match its angle and crop. Adding 3 or 4 images from different angles often helps with turns and profile views.
How is this different from Wan 2.2 Animate?
Wan 2.2 Animate replaces a character using 1 image and no prompt, at up to 720p. Kling 3.0 Omni accepts up to 4 images, requires a prompt, supports 1080p and offers four aspect ratio options.
Start Your First AI Video Swap
Upload a clip, add your subject images and describe the change. You see the credit cost before you generate, and new accounts start with 50 free credits.
Try Video SwapLearn more
What the AI Video Swap Model Can Do
Up to 4 subject images
Add 1 image for a quick test or up to 4 to show the new subject from several angles. A front view, a three-quarter view and a full-body or full-product shot give the model far more to work with than a single selfie.
Prompt-guided transformation
A text prompt is required. Use it to say who or what should be swapped, what to keep, and any detail the images cannot show, such as how a jacket should move or which hand holds the product.
Four framing options
Choose auto, 16:9 for YouTube and websites, 9:16 for Reels, TikTok and Shorts, or 1:1 for feeds and marketplace listings, all from the same upload.
720p or 1080p, audio on or off
Render a 720p draft to check the swap, then go to 1080p for delivery. The audio switch lets you output the result with sound or as a silent clip for your own soundtrack.
Clear credit cost before you start
The model starts at 264 credits. The studio shows the exact price for your settings before you press generate, and failed runs are refunded automatically.
What Is AI Video Swap?
AI video swap is a way to take a clip that already works, such as a dance, a product demo or a walk-through shot, and replace the subject in it without reshooting. On NeoMundo this runs on Kling 3.0 Omni, the multimodal Kling model from Kuaishou, in its transformation mode: it reads your source video, your subject images and your prompt together, then outputs a new clip in which your subject performs the original action.
This is different from motion transfer. With Motion Control you start from a still character and borrow movement from a reference video. With AI video swap you start from the finished video and change who or what appears in it, so framing, camera path and timing come from footage you already approved. That makes it useful when a shot is good but the person, outfit or product in it needs to change.
The inputs are simple: exactly 1 source video (MP4, MOV or WebM up to 100 MB), 1 to 4 subject images (JPG, PNG or WebP up to 10 MB each) and a written prompt. Settings are aspect ratio (auto, 16:9, 9:16, 1:1), resolution (720p or 1080p) and audio on or off. Videos usually take about 1 to 5 minutes depending on queue and settings.
Kling 3.0 Omni lives in the Motion & swap studio next to Motion Control and Wan 2.2 Animate. If you only need to replace person in video with a single image and no prompt, Wan 2.2 Animate is the lighter option; if you want several reference images and written control, use this model.
How to Use AI Video Swap in 4 Steps
Step 1: 1. Upload a clean source video
Pick a clip where the subject you want to replace is large in frame, well lit and not hidden behind other people or objects for long stretches. Steady footage with one clear main subject gives the most stable result.
Step 2: 2. Add 1 to 4 subject images
Use sharp photos of the new person or product on a simple background. Match the source where you can: if the video shows a full body, include a full-body image; if it shows a product in hand, include a shot at a similar angle.
Step 3: 3. Write the swap prompt
Name the target and the replacement, for example: "Replace the woman in the red coat with the person in the images, keep the street, lighting and camera move." Short, specific instructions beat long style lists.
Step 4: 4. Choose settings and generate
Set aspect ratio, 720p or 1080p and audio on or off, check the credit cost shown, then generate. Download the result, or run a new AI video swap with adjusted images or prompt.
AI Video Swap Use Cases
Product versions of one ad
Shoot one strong demo, then swap in each color or model of the product from your catalog photos instead of filming every variant.
Localized creator and UGC ads
Keep a proven script and camera flow while you replace person in video with a presenter from your own team or a creator who has agreed to appear, and export 9:16 for short-form feeds.
Outfit and styling tests
Try a new look on an existing runway, walking or try-on clip by adding images of the outfit and describing what should change.
Storyboards and pre-visualization
Drop your own character into reference footage to test casting, costume or mascot ideas before a real shoot.