Ballon d'Or AI Prompt: Start With One Clear Action

Kling or another image-to-video editor: a single stage shot

If your chosen editor offers image-to-video, upload the portrait using its image control and paste this visual direction. Choose a supported length and portrait format in that editor; writing “10 seconds” in prose does not enable an unsupported output length. This draft deliberately avoids a complicated walk and cut.

Use the supplied photo as the appearance reference for one consenting adult. A fictional football awards gala, warm stage lights, ordinary unnamed adult guests. The person wears [a black suit and tie / a black evening gown] and holds a plain golden sphere trophy at chest height with both hands. They look toward the camera, give a small proud smile and lift the trophy slightly. One continuous medium shot, slow gentle camera push-in. Keep the same face, hair, body proportions and outfit throughout. Natural hands, one stable trophy, no readable text, no logos, no famous people. No recognizable song or quoted dialogue.

Seedance: separate likeness from the reference performance

ByteDance describes multimodal referencing for Seedance. If the interface you use accepts an authorized reference video, state what comes from each input. The @Image1/@Video1 labels below are illustrative: attach the files through that interface and use its actual reference syntax. A name in a text box does not upload a file. Do not use a video you lack permission to process.

Appearance reference: @Image1, one consenting adult. Motion and camera reference: @Video1, an authorized fictional award sequence. Replace only the main winner with the adult from @Image1 wearing [a black suit and tie / a black evening gown]. Follow the reference performance, camera movement and shot transitions while keeping the winner’s face and body proportions consistent. Show ordinary unnamed adult guests and a plain golden sphere trophy. Keep the trophy shape and outfit stable. No captions, logos, famous people or recognizable music. Render a fictional entertainment clip, not real award footage.

Veo workflows: split a long sequence into two shots

Google’s Veo guidance separates subject, action, camera and atmosphere. For a gala sequence, try an audience shot and a stage shot separately, then edit them together. Confirm which Veo product or model your account can currently use: Google states the Veo 3.1 Preview endpoints in the Gemini API were shut down on October 22, 2026. We do not provide a supposedly current API command for those endpoints.

Shot A: one consenting adult from the supplied appearance reference, wearing [a black suit and tie / a black evening gown], seated among ordinary unnamed adults at a fictional gala. The winner looks surprised, stands and takes one step into the aisle. Warm gold light, medium shot, gentle tracking camera.

Shot B: the same adult, same face and outfit, already on a small gala stage, accepting a plain golden sphere trophy and smiling toward the camera. Medium close-up, restrained gold confetti, natural hands, stable trophy shape. No logos, readable text or famous people. Ambient applause only if the chosen model supports audio; no recognizable song.

Change one variable when the result fails

Face changes: use a sharper front-facing portrait and simplify the camera movement. Hand or trophy distortion: keep the ball at chest height rather than asking for a kiss, spin and overhead lift in one shot. Outfit drift: keep one outfit description in every shot.

Wrong timing: text alone cannot promise frame-for-frame matching. Use an editor with actual reference-video controls when you have authorized source material. If your goal is this site’s fixed ceremony sequence rather than a new creative scene, use the maker instead.

Make the fixed 10-second award sequence

What has and has not been tested

These are original prompt drafts written for this guide. We have not run them as paid generations or benchmarked the three providers. The site’s existing 5-second 480p sample used an earlier Mini model; it does not validate these prompts or the current 10-second workflow.

Before publishing, review likeness, hands, trophy shape, text artifacts and any soundtrack rights. Use photos only with the adult subject’s consent and disclose that the scene is AI-generated. A provider’s broad capabilities do not prove that every account, region or API wrapper exposes them.