About the Central Cee Doja AI Video Generator
The Central Cee Doja AI Video Generator puts a character of your choice in the lead role of the Doja video. You upload one image of your main character, and the AI rebuilds every shot with that person in place of the original performer while the moves, camera work, cuts and audio stay exactly as they were.
The clip also features friends, a crowd and dancers. If you want them recast too, add up to seven more images and the template assigns each one to a supporting person. Anyone you don't supply a reference for stays as they appear in the source, so you can recast one person or eight.
How to make a Central Cee Doja AI video
- Upload your main character (Image 1). Image 1 is required. It becomes the lead performer in every shot, so use a clear image of one person with the face, hairstyle and full outfit visible.
- Add supporting characters if you want them (Images 2–8). You can add up to seven optional images. They are assigned to supporting people in the order those people first appear in the clip you pick; when several appear together, they are assigned from left to right. Remove an image and the ones after it move up to fill the gap.
- Pick a resolution. Every video is 15 seconds long. Choose 480p or 720p; the credit cost is shown next to the Generate button before anything is submitted.
- Generate, preview and download. Press Generate. You are taken to your generation, which shows its progress while the video renders. When it finishes you can play it on the page and download the MP4. If the render fails, the error is shown and the credits are returned automatically.
How the template works
The template uses motion transfer. The AI takes the performance from the source clip (every pose, gesture, camera move and cut) and applies it to the people in your images. Rather than pasting a face onto the original performer, it rebuilds the person from your reference, including their face, hairstyle, body proportions, clothing and accessories.
Everything that isn't a person you've recast is kept from the source: the tower blocks, streets, lighting, props, shot timing and the synchronised audio. The template doesn't add people, scenes or objects.
Uploading character references
- One person per image. Each reference should show a single person. Group photos make it unclear who you mean.
- Show what you want on screen. The model copies clothing and accessories from the reference, so dress your character the way they should look in the video.
- Full body helps. A sharp, evenly lit image with the whole body in frame works best. A character sheet with front and back views gives the model even more to work from.
- Order matters. Image 1 is always the main character. Images 2 to 8 follow the order in which supporting people first appear in the 15-second clip, and people who appear together are assigned left to right.
- Formats. JPG, PNG or WEBP. Very large photos or other formats are converted in your browser before upload.
- Consent. Only upload images of people who have agreed to appear, or characters you have the rights to use.
How length and resolution affect the result
Every video follows the same 15-second source clip, so supporting references always land on the same people in the same order. The clip is a re-cut of the original video that keeps the performer's solo shots, portraits and dance moments.
480p renders faster and costs fewer credits, which makes it a good way to check your casting. 720p gives a sharper result for posting. The price table is below.
Previewing and downloading your video
The source previews on this page show the original reference clips so you can see who appears and in what order before you assign supporting characters. They are source footage, not generated results.
After you press Generate, your video appears in your generations with a live progress status. It's safe to leave the page while it renders. When it's done you can watch it in the browser and download the MP4 with the original audio.
Resolution and credit cost
| Duration | 480p | 720p |
|---|---|---|
| 15s | 4,055 credits | 8,683 credits |
Credits are charged once when the render starts and refunded automatically if the generation fails or is blocked by moderation.
Frequently asked questions
Which images are required, and which are optional?
Only Image 1, your main character, is required. You can generate with that one image alone. Images 2 to 8 are optional supporting characters; add as many as you like, up to seven.
How are main and supporting characters assigned?
Image 1 is always the main performer in every shot and is never used for anyone else. Images 2 to 8 are assigned to distinct supporting people in the order they first appear in the clip you select. People who appear together for the first time are assigned from left to right. Each assignment is kept for the whole video, and one reference is never applied to several people.
What happens to people I don't upload a reference for?
They stay as they appear in the source clip. If you upload more supporting references than there are people in the clip, the extra references are ignored rather than turned into new people or scenes.
Can a man take Rosé's role, or a woman Bruno Mars's?
Yes. Each person keeps their own gender, face and build from their photo, whichever role they take, and wears an outfit similar to that role's, fitted to them.
How long is the video?
Every video is 15 seconds long and follows the same source clip, with the original audio. You choose the resolution: 480p or 720p.
Should I choose 480p or 720p?
480p is cheaper and faster, which is handy for testing a casting. 720p is sharper and better for posting. Both keep the landscape framing of the source clip.
Is the original audio kept?
Yes. The synchronised audio from the source clip is preserved, and the performance stays in time with it.
Why does a face or outfit sometimes look inconsistent?
In crowded shots, fast movement, motion blur, heavy fisheye framing or when people overlap, the model can occasionally blend details or place a reference on the wrong person for a few frames. Clear full-body references of one person each, and fewer supporting characters, reduce this. Try 480p first to check a casting before rendering the sharper 720p version.
Is this an official Central Cee template?
No. This is an independent AI template. It is not affiliated with, endorsed by or sponsored by Central Cee or anyone involved in the original video.
What happens if my generation fails?
The error is shown on your generation, and the credits for that render are returned to your balance automatically.
Discover more templates in the trends gallery or explore AI video effects.
The Central Cee Doja AI Video Generator is an independent AI template. It has no official affiliation with, and is not endorsed or sponsored by, Central Cee or the creators of the original video. The artist's name is used only to describe the source clip. Only upload images you have the right to use.