Skip to main content

The prompt

Lamine Yamal rabona prompt: two prompts to copy, and when to skip them

Plenty of people asked for the words behind the semi-rabona remakes. Here they are: two prompts you can paste into a general image-to-video tool with your own photo. One is short. One is shot by shot. Both describe the move and the stadium, not the player. Below them we say plainly what prompts tend to get wrong, so you can decide whether to write it yourself or skip straight to the maker.

Updated Oct 18, 2026: a short prompt, a detailed shot-by-shot prompt, a table of what a prompt has to say, and the hard parts.

1Your photo

Your photo is deleted from our storage after 30 days and is never used to identify anyone. How we handle your photo

Standard · 8s · 9:16 · 720p · with sound

Your first video $5.99 · then $9.99 a video (10 credits), or 35 for $99

No subscription needed · Ready in about 3 minutes · Failed renders are never charged

Photo in: an AI-made portrait of a man in an olive hoodie and grey joggers. Not a real person. First take.

The short answer

A good Lamine Yamal rabona prompt never asks for Lamine Yamal. It describes the move from his Champions League match against Galatasaray on 14 Oct 2026, with "the person in the photo" on the ball: a run down the right wing under floodlights, the left foot planted, the right leg swung behind the standing leg as if to hit a rabona, no strike, a defender in a red/orange shirt sliding past, then a drag inside and a carry toward the box. Seen from a high, wide TV camera, 8 seconds, vertical, crowd sound. Copy one of the two prompts below, or skip the prompt and make yours here from one photo.

What's new on this page

  • 2026-10-18Page published with two copyable prompts: a short one and a detailed shot-by-shot one.
  • 2026-10-18Added a table of what the prompt has to say and why, plus the four things prompts most often get wrong.

One rule

Before you copy: describe the move, not the player

The moment everyone is remaking is Lamine Yamal's "no rabona": on the right wing he shapes up for a rabona, never kicks, the defender slides past, and he cuts inside. The AI wave around it moved fast. We counted 185 AI remakes between 14 and 18 Oct 2026, with 16.5M plays between them, and 67 of those went up on 16 Oct alone.

Not one of those remakes needs his face. The point is to put yourself, a friend or your dad on the ball. So the prompts here say "the person in the photo" and describe what happens, where the camera is and what you hear. They never name a real player, a club or a competition, and they ask for no crests or kits. That keeps the result yours, and it keeps the tool from refusing the request.

Short prompt

Prompt 1: the short one

Use this when a tool has a small text box or does better with fewer instructions. It packs in the four things that make the clip recognisable: the high wide camera, the leg behind the leg, the missed slide, and the carry toward goal.

Short prompt (copy)
Vertical 9:16 video, 8 seconds, realistic football match at night. Wide high TV broadcast camera looking down at a floodlit green pitch. The person in the photo, wearing the clothes from the photo, runs down the right wing with the ball, plants the left foot and swings the right leg behind the standing leg as if to hit a rabona, but does not kick. A defender in a plain red-orange shirt slides in and misses. The person drags the ball inside and runs on toward the penalty box. Keep the face from the photo. Stadium crowd sound, no commentary. No text, no logos, no crests, no scoreboard.

If a tool caps the length, cut from the end, not the start. The camera and the leg swing matter most; the list of things to leave out matters least.

Detailed prompt

Prompt 2: the detailed one, shot by shot

This one follows the beats of the original second by second, stretched to an 8-second clip. The template clip people remake is about 7.5 seconds, so the last beat runs a little longer. Use it when a tool accepts long prompts and follows timing.

Detailed shot-by-shot prompt (copy)
Vertical 9:16 video, 8 seconds, realistic, one continuous shot, no cuts.

Scene: a large football stadium at night, floodlights on, stands dark and full in the background. Green grass with white lines. The camera is the TV main camera: high in the stand, looking down about 30 degrees, long lens, wide framing. The player is small in frame and seen full body. The camera pans slowly to the right to follow the ball.

Player: the person in the photo. Keep their face, hair, skin tone and the clothes from the photo exactly. They play in those clothes.

Defenders: two outfield players in plain red-orange shirts with no badges, no numbers, no sponsors. Generic faces.

0-1 s: the person runs onto the ball from the left side of the frame, ball at their feet.
1-2 s: they carry the ball forward along the right touchline at speed. A defender closes in.
2-3 s: they plant the left foot next to the ball and swing the right leg wide and BEHIND the standing left leg, as if to strike a rabona. The legs cross clearly.
3-4 s: they do NOT strike the ball. The defender has already committed and slides in feet first, missing the ball and sliding past.
4-5 s: they drag the ball inside with the outside of the right foot. A second defender lunges and misses.
5-6.5 s: they carry on toward the penalty box, past the defender on the grass. The goal post enters the frame on the right.
6.5-8 s: near the corner of the box, goal post and net in view. The clip ends on the run.

Sound: stadium crowd only, rising when the defender misses. No commentary, no music.
Do not show: text, captions, a score or clock, channel logos, club crests, real kits, real players.

The capital letters on BEHIND and NOT are there on purpose. Those are the two words tools ignore most often, and a little shouting helps some of them listen. For the exact clip our maker recreates, shot by shot and camera to sound, see the football skill AI video page.

The parts

What the prompt has to say, and why it matters

Whether you use our prompts or write your own, these are the parts that decide if people recognise the meme. Leave one out and the clip slides toward "someone doing a trick".

Each part of a semi-rabona prompt
What the prompt has to sayWhy it matters
High wide TV camera, about 30 degrees downIt is the view everyone saw. Without it, it stops looking like a match.
Floodlit night stadium, green pitch, white linesThe look of the original. Daylight or a park pitch reads as training.
Plant the left foot, right leg swings BEHIND the standing legThis is the rabona wind-up, the frame people know. A normal swing is just a kick.
Does NOT strike the ballThe whole joke is the fake. If the player kicks, the meme is gone.
Defender commits, slides in and missesThe payoff. The fake only lands if someone falls for it.
Drag inside with the outside of the foot, carry toward the boxFinishes the move and gives the clip somewhere to go.
The person in the photo, same face and clothesIt has to be you. Saying it plainly helps the tool stick to your photo.
Player small, full body, camera pans to followKeeps the feet in frame, so the leg cross is visible.
Crowd sound, no commentaryFeels like the stadium and leaves the words to your caption.
No text, logos, crests, scoreboard or real playersKeeps the clip yours and keeps tools from refusing the request.
Vertical 9:16, 8 seconds, one shotFits TikTok, Reels and Shorts, and stops the tool cutting away.

What to leave out of your prompt

  • Real names: the player, the clubs, the competition. Describe, don't name.
  • Kit details that copy a real team, like stripes in club colours or a sponsor.
  • Style words that fight the broadcast look, like "cinematic close-up" or "slow motion".
  • Camera moves that break the single wide shot, like "zoom in on the face".

Skip it

Or skip it: what is hard to get right by prompt

A prompt is a request, not a guarantee. General video tools are built to do anything, so they are not tuned for one football move from one camera angle. These are the four places where a typed prompt most often falls short.

1. The leg behind the standing leg

Crossing one leg behind the other is an unusual body shape. Tools tend to swap it for a normal kick, swing the leg in front, or blur both legs into one where they overlap. It is also the single frame that makes the clip readable, so getting it wrong ruins the whole thing.

2. The defender's timing

The defender has to commit to the slide at the moment of the fake, not before and not after. Too early and it looks like a foul; too late and the player just runs past someone standing still. Timing words in a prompt help, but many tools treat them loosely.

3. Keeping the face in a wide shot

The broadcast camera makes the player small. At that size, plenty of tools let the face drift into someone generic, or they zoom in to save it and lose the broadcast look. Keeping both, a wide shot and your face, is the trade-off a prompt can't settle for you.

4. Sound

Many tools give you a silent clip, or music, or a voice that sounds like commentary in no real language. A clip with no crowd feels empty on TikTok. Adding sound afterwards means another app and another export.

Our maker sidesteps all four, because it does not ask a tool to invent the move from words. We recreate the trend clip itself with you on the ball, and no prompt is needed. It is the same 8-second clip every time; only the person on the ball changes. The run onto the ball, the carry, the rabona wind-up with the leg behind the standing leg, the missing strike, the defender in red/orange #17 sliding past, the cut inside and the carry toward goal all happen exactly as in the viral clip, from the same wide, high camera, with stadium sound. You upload one photo, and the player on the ball gets your face, your hair and the clothes from that photo. Vertical 9:16, 720p, about 3 minutes.

One honest limit: the face is small in frame. It is a wide match camera, so think highlight reel, not close-up. And it is one clip, not a menu: you cannot pick another scene, angle or outfit. The video is AI-generated, not real footage.

It runs in a phone browser, with no app, no CapCut and no TikTok account. A video is 10 credits. Single is $9.99; the Duo Pack is $29 for 5 videos, the Crew Pack $99 for 35, and pack credits never expire. No subscription needed, and failed renders are never charged: the credits come back on their own. See pricing for every option.

Photos must be of you or of someone who said yes: one person, alone, full body or waist-up, face clear and front-facing, in good light, with no sunglasses or hat over the face. Group photos, side profiles, blurry or dark photos and pets don't work. No celebrities, footballers or public figures. Anyone under 18 needs a parent or guardian to agree and do the upload. Photos are deleted after 30 days. More about us on the about page.

Compare

Prompt or maker: a quick comparison

Your own promptOur maker
What you give itA photo and a block of textOne photo, no prompt
What you getA new clip the tool invents from your wordsThe trend clip, recreated with you on the ball
The leg behind the legHit or miss, often several triesThe same as the trend clip, every time
Camera, pitch and defenderDepends on how the tool reads itThe trend clip's wide camera and defender #17, every time
Timing of the slideTreated loosely by many toolsIdentical in every take
Your face in a wide shotOften driftsFace, hair and clothes from the photo; small, as in a match shot
SoundDepends on the toolStadium crowd, no commentary
Failed attemptUsually still costs youNever charged

If you enjoy tinkering, the prompts above are a good start and they are yours to keep, and a tool may give you a version of your own. If you just want the clip people know, with you on the ball, make yours here. And if you are curious what the move actually is, step by step, read the semi-rabona meaning.

Asked in the comments

Questions about the rabona prompt

Is there a prompt for the Lamine Yamal rabona AI video?

Yes, two of them are on this page: a short one and a detailed shot-by-shot one. Both describe the semi-rabona feint with "the person in the photo" on the ball, so they work with your own photo in a general image-to-video tool.

Can I put Lamine Yamal's name in the prompt?

We would not. Asking a tool to show a real player can get the request refused, and it is not what the meme is about anyway: people remake the move with themselves on the ball. Our prompts describe the move and the scene, never the player.

Are these prompts free to use?

Yes, copy them and use them wherever you like. The tool you paste them into may charge for its own renders; that is between you and that tool.

Why does my AI video not do the leg behind the leg?

It is the hardest part. Most tools default to a normal kick, or the legs blur together where they cross. Saying "behind the standing leg" and "the legs cross clearly" helps, but it often takes several tries. That is the main reason people skip the prompt and use a maker built for this one move.

Will the prompt keep my face?

Sometimes. In a wide broadcast shot the player is small, and many tools drift away from the photo at that size. A clear front-facing photo and the line "keep their face exactly" help. Our maker works differently: it recreates the trend clip with you on the ball and keeps your face, hair and clothes from the photo. It is still a wide match camera, so think highlight reel, not close-up.

Can I just use your maker instead of a prompt?

Yes, and no prompt is needed. Upload one photo and we recreate the trend clip with that person on the ball: the same 8-second clip every time; only the person on the ball changes. Vertical 9:16, 720p, stadium sound, about 3 minutes, 10 credits. Single is $9.99. Failed renders are never charged.

Does your maker show Lamine Yamal?

No. It shows the person in your photo on the ball, never Lamine Yamal or any real player. The pitch, the defender in the red/orange number 17 shirt, the camera and the moves stay as they are in the trend clip. The video is AI-generated, not a real recording.

Rather not write a prompt?

Upload one photo and get an 8-second vertical video of that person faking the rabona under the floodlights, with crowd sound, in about 3 minutes.