Skip to main content

AI prompt

Ozil goal AI prompt

Our generator needs no prompt: one photo in, a 15-second clip out. But if you want to try the Ozil goal scene in another image-to-video tool, the builder below writes a clean prompt for you. Pick a few options, copy, paste.
Make my video

One photo · 15s · from $5.70 a video

Prompt builder

Build your Ozil goal prompt

Every choice rewrites the prompt live. The text stays plain English with no special characters, so it pastes cleanly into any tool. Nothing you type leaves your browser.

Clip length
Prompt style
What they wear
Setting
Camera
Ending
Vertical 9:16 video, 10 seconds, one continuous shot, no text on screen.
Main person: the adult in the reference photo. Keep their face, hair, skin tone and body shape the same in every frame. They wear exactly the clothes shown in the reference photo, unchanged. Not a football kit.
Other players: one goalkeeper in a plain orange kit and one defender in a plain dark blue kit, ordinary faces, no names, numbers, badges or logos.
A large football stadium at night under bright floodlights, green turf, a packed crowd softly out of focus.
Camera: TV broadcast angle from the side of the pitch, slightly above, long lens, smooth pan that keeps the main person in frame, full body visible.
0 to 2s: a goalkeeper rushes out and throws himself into a full flying dive, body flat in the air, and misses the ball as the main person nudges it past him.
2 to 5.5s: the ball pops up and drops onto the main person, who cushions it off the chest and lets it bounce back down to their feet.
5.5 to 10s: a defender slides in hard along the grass and misses; the main person runs on and calmly slots the ball into the empty net.
Sound: stadium crowd noise rising to a roar, the thud of the ball. No commentary, no music with lyrics.
Avoid: football kit on the main person, face changes, extra limbs, text, subtitles, scoreboards, logos, club badges, watermarks, the likeness of any real footballer.
249 words · plain text, no real names

Using it

How to use the prompt in another tool

  1. Pick one photo of one adult who agreed to it. Face clear, good light, and ideally the whole outfit in view. Do not use footballers, celebrities or children.
  2. Set the tool to vertical 9:16 and to the same length you chose in the builder. If the two disagree, the last beat is usually the one that gets cut, and the last beat is the payoff.
  3. Paste the prompt and render the cheapest version first. Short and low resolution is enough to see whether the face holds and whether the moves happen in order.
  4. Change one thing at a time. If the dive is missing, do not also change the camera and the outfit. One change per render tells you what actually helped.
  5. Label it as AI when you post. TikTok, Instagram and YouTube all have a setting for AI-generated content.

Why one-shot prompts often miss beats

The trend packs three distinct actions into a short clip: a goalkeeper diving flat through the air, a ball bouncing off the chest, and a defender sliding in and missing. Image-to-video models are generally best at one clear action per shot. Ask for three in a row and the usual result is that one of them happens well, one happens half-heartedly and one quietly disappears.

Timestamps help because they give the model an order, but they are a hint, not a schedule. Longer clips also give the face more time to drift away from your photo. That is why the builder has a Three shots option: each beat becomes its own short prompt with the same person, outfit, setting and camera lines, so you can render them separately and join the best take of each in any editor.

Our own generator handles this differently: the timing and the moves come from the trend clip itself, and your photo only supplies the person. That is why it needs no prompt. If you want to compare the routes side by side, see the AI template guide or the CapCut comparison.

Line by line

What each part of the prompt does

If you edit the prompt by hand, this is what each block is for. Cut a line only if you know what you are giving up.

BlockWhat it saysWhy it is there
Format lineVertical 9:16, length, one continuous shotTells the tool the frame and stops it inventing cuts or captions.
Main personKeep face, hair, skin tone and body the sameIdentity drift is the most common failure. Saying it up front helps.
OutfitClothes from the photo, not a football kitThe joke is an ordinary person among players. Kit kills it.
OpponentsPlain kits, no names, numbers or badgesKeeps real clubs and real players out of the frame.
Setting and cameraStadium, light, angle, full body visibleA side-on, full-body view is what makes the moves readable.
Timed beatsDive, chest bounce, slide and finishTimes give the model an order to follow, even if it does not hit them exactly.
Sound and avoidCrowd noise, no lyrics; a list of things to leave outCuts the usual extras: text, logos, watermarks, extra limbs.

When it breaks

Fixes for common problems

AI video is unpredictable, so expect a few tries. These are general fixes that apply to most image-to-video tools.

ProblemTry this
Your face changes halfway throughUse a sharper, front-facing photo in even light. Shorten the clip, or switch to Three shots so each render is brief.
You end up in a football shirtPick an outfit preset or type a detail like "brown cardigan". Use a photo that shows the outfit clearly.
The keeper only slides on the grassRender the dive as its own shot. Many models default to a slide because it is the more common move.
The ball never leaves the groundRender beat two on its own and keep the camera side-on so the bounce is visible.
Text, logos or a scoreboard appearKeep the Avoid line. If it persists, try the high and wide camera, which shows fewer boards close up.
The clip stops before the finishMake sure the length set in the tool matches the length in the prompt.

Or skip the prompt

Prompting is fun if you like tinkering. If you just want the clip, our generator takes one photo and returns a 15-second vertical MP4 with the trend’s sound in about 5 to 8 minutes. One video is $8.90, three are $19.90 and seven are $39.90, paid once with no subscription. If a render fails, the video goes back to your balance automatically.

Asked in the comments

Ozil AI prompt: questions

Do I need a prompt to make the Ozil goal video on this site?

No. Our generator already has the scene, the players, the camera and the timing built in. You add one photo and press Make my video. The builder on this page is for people who want to try the scene in a different image-to-video tool.

Which AI video tool should I paste the prompt into?

Any tool that accepts one reference photo plus a text prompt and can export vertical video. We do not test or rank third-party tools, and their behaviour changes often, so try a short, cheap render first and adjust from there.

Why does my result skip the keeper dive or the chest bounce?

Packing three actions into one short clip is hard for most models: they tend to keep the first or the most common action and blur the rest. Switch the builder to Three shots, render each beat on its own and join them in an editor.

Why do I end up in a football shirt?

A stadium scene pulls the model towards kit. Keep the line about the clothes from the photo, add a short outfit detail, use a photo where the outfit is clearly visible, and keep the opponents in colours that differ from what you wear.

Should I write Mesut Ozil’s name in the prompt?

No. Naming a real player invites the model to draw his face or a real kit, and many tools block real names anyway. The builder describes the moves instead and asks the model to avoid any real footballer’s likeness.

Can I use and change the prompt?

Yes. Copy it, change it and use it wherever you like. Rendering in another tool may cost money there. If you would rather skip the prompting, a video on this site is $8.90, or from $5.70 each in a pack.

Skip the prompt. Just add a photo.

Make my video

15s · 9:16 · from $5.70 a video · a failed render costs nothing