Copyable prompt
Create one cinematic, photorealistic 8-second vertical 9:16 fictional football highlight. Use the uploaded portrait only as the person and identity reference: preserve their recognizable face, hair, skin tone, and build. Use the permitted action reference video only as a guide for football movement and timing; do not treat either input as a first or last frame. Reimagine the person as a fictional footballer in a plain, unbranded kit on a fictional floodlit European pitch. Keep the action in one continuous, readable camera move: 0–1.5s, the player takes two quick controlled touches toward goal; 1.5–3.5s, the goalkeeper rushes out and the player lifts a precise chip over the dive; 3.5–6s, the ball drops, bounces once, and rolls into the open net; 6–8s, the player turns toward camera for a joyful, natural celebration as the stadium crowd erupts. Make the chip, ball arc, bounce, and goal clearly visible; use realistic foot-to-ball contact, grounded anatomy, and smooth motion. Add natural stadium ambience, a rising crowd roar, a soft kick, and a clean net sound; no dialogue, lyrics, or music. Frame the player and ball safely inside the vertical image. Do not add text, subtitles, watermarks, real-player likenesses, club crests, sponsor marks, or branded kits. This is a fictional AI football scene, not match footage.
Why the prompt uses one action chain
A goalkeeper rush, chip, bounce, defender dodge, and finish already ask a short clip to show several changes. Keeping them in one continuous action gives the model a clearer order than adding multiple camera cuts, extra players, crowd details, and text overlays.
What to customize first
Keep the subject and action order fixed, then change one visual choice at a time: the stadium lighting, camera movement, or kit color. Avoid adding real team marks, sponsor logos, broadcast graphics, or a real player’s face. If an action disappears, simplify the sequence or use a longer clip setting where available.
Ozil AI does not require prompt writing
The copyable prompt is for workflows where you can edit the prompt yourself. Ozil Goal AI uses a preset scene prompt and asks for an adult portrait plus a permitted action reference clip, so you can create the fixed 8-second, vertical 9:16, 720p version without composing text.
Set expectations for the result
The prompt asks for a fictional scene inspired by the broad action sequence. It cannot restore the original broadcast, guarantee an exact face or movement, or ensure that every beat appears in the final clip.
Step-by-step
- 1. Start with one clear reference photo
Use a permitted photo of an adult with the face and upper body visible. Avoid a crowded image with several possible subjects.
- 2. Keep the action order simple
Describe the keeper rush, chip, bounce, defender recovery, and finish as one continuous play. Remove extra actions if the model loses the sequence.
- 3. Add camera direction and exclusions
Ask for a readable camera path and realistic ball movement, then exclude real likenesses, team branding, broadcast marks, and on-screen text.
- 4. Review and adjust one detail
Check the result for the action beat you wanted. Change one phrase at a time so you can tell which instruction helped.
Frequently asked questions
Can this prompt recreate the exact Özil broadcast clip?+
No. It requests a fictional football scene inspired by the sequence, not the original match footage or an exact recreation.
Can I use the prompt with one portrait?+
Yes, in an image-to-video workflow that accepts a reference photo. Results vary by model and settings.
Do I need this prompt to use Ozil AI?+
No. Ozil Goal AI uses a preset prompt, so its portrait-and-action flow does not ask you to write or paste one.
Why might the finish or bounce be missing?+
The clip is short and the model may simplify a multi-step action. Keep one subject and one camera path, or choose a longer duration when your workflow supports it.