An Ozil Goal AI prompt is an instruction for a synthetic football scene: a recognizable person moves through a composed finish, with the ball, goalkeeper, and defenders contributing to a short visual story. It is useful to understand those instructions before choosing a tool. It is also useful to know which parts of a result a prompt cannot control reliably.
On Ozil Goals AI, you do not need to supply that text. Our generator uses a fixed motion preset, a reference video, and your uploaded photo. The preset is designed around one scene rather than a list of selectable sports actions. This guide explains that approach and provides a separate, clearly marked example for people learning how general video prompts work.
What a football motion prompt needs to describe
A useful instruction establishes the subject, the action, and the camera's job. These three elements can easily compete. A close-up helps viewers recognize a face, but it can hide the ball. A wide view makes the sequence readable, but it reduces the amount of facial detail visible on a phone screen.
For a football finish, continuity matters more than adding cinematic adjectives. The same person should remain visible across the shot. Their outfit should stay recognizable. The ball should occupy a plausible position relative to the feet. The goalkeeper and defenders should not appear or vanish without a reason. These are goals for the generation, rather than guarantees that a model will satisfy every instruction.
A reference video provides information that is awkward to communicate in prose alone, such as the spacing between a lifted ball, a pause, and a final touch. That is why copying a paragraph from the internet is not equivalent to using the same motion preset. A different model, source image, or reference can produce a substantially different result from identical words.
How the fixed preset works on this site
You provide one photo of yourself or another adult who has given permission. A full-body photo is the strongest starting point because it shows the face, clothing, legs, and shoes together. A knee-up photo may be usable, but leaves the model to infer more of the lower body. Our photo preparation guide covers the trade-offs in detail.
The service combines that photo with the fixed football scene and submits the job asynchronously. The configured output is a 15-second, 720p, vertical 9:16 MP4. There is no customer-facing prompt editor, alternate goal selector, or model settings panel. Read the template explanation if you want to compare the scene with the demo before purchasing credits.
The API receives a silent version of the motion reference. Generation is requested without audio, and the saved template soundtrack is attached after the visual result comes back. This separates visual generation from the soundtrack. It does not establish a music license for every destination where a clip might be published.
An illustrative prompt for a separate general video tool
The example below is educational text for an external model that accepts a photo, text instructions, and an authorized motion reference. It is not a field you can paste into our form, not our internal production prompt, and not a promise that a third-party service supports these inputs.
Create a fictional vertical football video featuring the consenting adult in the supplied photo. Use the permitted motion reference to guide the order and rhythm of the finish. Keep one main subject, recognizable clothing, a coherent football pitch, and a ball that remains visible through the action. Show a composed approach, a lifted touch beyond the goalkeeper, a brief adjustment around the defenders, and a final shot. Keep the camera wide enough to read the movement without unnecessary cuts. Aim for consistent identity and body proportions. Do not introduce commentary, text overlays, official branding, or generated music.
Treat that paragraph as a brief, not a magic formula. Check the external tool's supported inputs and content rules before using it. If it only accepts text, describing a reference does not give it access to that reference. If it has a short duration limit, asking for a longer sequence may compress the action or produce an incomplete finish.
Prepare the image before adding more instructions
When a face is blurry or covered, repeatedly adding “preserve identity” does not recover missing visual information. Choose a photo where the person's features can be seen at normal viewing size. Avoid sunglasses, heavy beauty filters, objects covering the torso, and group photos with several plausible subjects.
Try a straightforward standing pose with natural light. Leave space around the body rather than cropping tightly into the shoulders or feet. Distinctive but simple clothing gives the model useful continuity cues. Tiny logos and complex patterns can distort during motion, so do not depend on a readable shirt number to carry the joke.
Use a photo you are permitted to animate. Do not upload children's pictures or celebrity photos. A fictional entertainment scene is easier to understand when the person being represented knows the purpose and agrees to it.
Judge the sequence rather than one attractive frame
Watch the beginning, the fastest movement, and the final shot. A result can start with a convincing face and still change identity halfway through. Look for extra limbs, sliding feet, a ball that disappears, clothing changes, or a sudden camera movement that hides the important action.
Text instructions also cannot guarantee an exact recreation of a professional player's technique. The output is a fan-made interpretation, with variations between generations. It should not be presented as real match footage or an official endorsement. For a broader walkthrough, see how to make an Ozil Goal AI video.
Understand the cost before experimenting
Each generation costs 8 credits. The available one-time packs are $5.99 for 8 credits, $23.99 for 32 credits, and $59.99 for 88 credits. Credits never expire. There are no subscriptions, free generations, free trial, or signup credits. The pricing page provides the purchase options, and our free-tool guide explains the distinction between viewing a demo and paying for a generation.
If a generation fails, the credits for that failed job are returned once. That is a credit return, not an automatic cash refund of a pack purchase. Completed media is kept for 30 days, so save the MP4 from Your goals when you are ready to keep or edit it. Better photo selection is a more practical first step than purchasing repeated attempts in the hope that extra prompt wording will make every detail exact.