How to make a Celebrity Kitchen AI video.
Use two adult portraits to cast the coffee maker and milk pourer in a 15-second kitchen skit. This walkthrough covers the fixed coffee-fail generator, from choosing photos to downloading your video and adding finishing touches.
Updated
What you will make
A celebrity kitchen AI video is a short comedy scene: one person tries to make coffee while a second person handles the milk, and the counter ends up a mess. In this studio, your photos identify the two roles. The reference scene sets the action; you do not need to film the skit yourself.
The result is a generated video, so faces, movement and small scene details can vary. The home page labels our actual generated example separately from the original scene and third-party references.
Choose two photos and assign the roles
Open the generator and add one clear portrait to each photo slot. The two people do not need to appear in the same picture. Both roles need a photo, even if you only want to change one character.
- Coffee maker: the person on the right, in the foreground, working with the coffee.
- Milk pourer: the person on the left, in the background, handling the kettle and milk.
Use JPG, PNG or WebP files up to 10 MB each. Choose front-facing portraits with even lighting, one visible adult face per image, and permission to use the photos. You can replace or remove either selection before generating.
The two photos behind our real video example
These are the two AI-generated adult portraits used to make this studio’s coffee-fail example. They are fictional test identities, not customer photos or celebrity portraits.
Compare these portraits with the generated clip. The recipe keeps the kitchen action and clothing from the reference scene; the photos identify the two people. This is one observed result, and likeness, movement and details can vary in a new generation.
Celebrity Kitchen AI video prompt and reference setup
Our generator applies this fixed reference-editing prompt. In the text below, @video_1 is the kitchen reference clip, @image_1 is the coffee-maker portrait, and @image_2 is the milk-pourer portrait.
Edit @video_1. Replace the person making coffee on the right in the foreground, wearing a blue-gray checkered shirt with the adult in @image_1. Replace the person pouring milk on the left in the background, wearing a black shirt with the adult in @image_2. Keep these two identities distinct and consistent through every frame, including the reactions. Preserve the reference kitchen, camera position, framing, body motions, cup, milk pouring, spill and comic timing. Preserve both characters' clothing from the reference video. Make both replacement faces naturally fit the original lighting and head movement. Keep exactly two people. Do not add text, logos, captions or watermarks. Generate natural kitchen sounds and reactions; no music and no spoken celebrity names. Output the complete 15-second vertical scene.The prompt is only one part of the setup: the reference video supplies the kitchen, action and comic timing. Pasting this text into an unrelated text-to-video tool will not reproduce the same setup. On this site, you choose the two photos; there is no prompt box or custom-scene control.
The prompt requests generated kitchen sounds without music. The public showcase has the original scene audio added afterwards; a new generated download does not automatically include that soundtrack.
Sign in and generate the kitchen video
Make sure your account has credits, then press Generate. Sign in if prompted; closing the sign-in dialog keeps your selected photos on the page. One generation uses one credit. The scene, duration and vertical format are already set, so there is no prompt to write or model to choose.
Selecting a photo creates a preview and a local draft. Pressing Generate starts the upload and generation process after sign-in. Your uploaded photos are sent for video generation. Read the photo privacy policy for processing and file-access details.
Check the current purchase status in Pricing before planning a generation. If purchases are paused, you can still watch the examples and prepare your photos; existing credits remain available.
Follow progress and download the MP4
The Your scenes area shows the actual job status. While a scene is in progress, let that job finish or check its status before submitting another. If the status cannot refresh, use Refresh status rather than assuming the job failed.
When the video is ready, play it from start to finish, then choose Download video. The output is a 15-second, 720p MP4 in 9:16 vertical format. Check both faces, their roles, hands, the mug and the milk pour before sharing.
A failed generation and a completed video you dislike are different outcomes. The job and account show the credit settlement; cancellation after processing starts can still incur generation costs.
Edit the downloaded video and add music
Celebrity Kitchen AI video editing starts here with choosing the two character photos for the fixed skit. After download, use a video editor for captions, cuts or music. This studio has no editing timeline or soundtrack selector.
The homepage and watch-page examples already include the original scene’s audio, added after generation. Unmute either example to hear it. New videos use model-generated audio, so their soundtrack can differ from these examples. Listen to your downloaded MP4 before posting. To use the original music or a different soundtrack, add audio you have the right to use in your editor, check its volume, and review the exported video.
Keep the vertical frame when your destination needs 9:16. Exporting a 720p clip at a larger size does not establish that extra facial detail was captured. Keep your original download so you can edit another version without generating the scene again.
Ready for your coffee fail?
Choose the coffee maker and milk pourer in the generator. Watch the real example before starting.
Choose your two photos See video credit packs