Street Robot Fight VIDEO Breakdown (Prompts Included)
AI Generated Video
Recreating this Street Robot Fight video is straightforward with Pollo AI. Follow the workflow below, prepare the same references, and paste the prompts to rebuild the viral AI video style quickly.
Text-to-Video Workflow
Open pollo.ai and select Video Generation, then choose the Text-to-Video mode. Paste your full video prompt, set your preferred video duration and aspect ratio. Select any one of these models: Seedance 2.5, Seedance 2.0, MiniMax H3 or Wan 3.0. Click generate once all settings are confirmed.
Video Generation Screen
Video Prompt
Create a 15-second vertical 9:16 ultra-realistic viral street video filmed casually on a smartphone from across a city street in daytime. ONE CONTINUOUS SHOT, no cuts. The footage should feel like an authentic unexpected real-world incident captured by a bystander, not a polished science-fiction movie. Handheld phone camera, moderate telephoto zoom, imperfect tracking, small framing corrections, realistic autofocus, natural daylight, ordinary traffic and street ambience. The main robot is a compact realistic humanoid service robot approximately 1.2 meters tall, with a smooth white hard-shell torso and limbs, black mechanical joints, articulated hands, slim mechanical legs and a glossy black featureless oval face visor. It must maintain exactly the same mechanical design, proportions and scale throughout. It moves using believable powered robotic joints but is surprisingly agile.
0-1s: Camera points across a normal downtown street toward the opposite sidewalk. Four young adult men are casually standing in a loose group outside a storefront, talking to one another. They wear ordinary hoodies, dark pants and sneakers. A gray SUV drives very close across the foreground from left to right, briefly blocking most of the scene like a completely natural traffic obstruction.
1-3.8s: As the SUV clears frame, reveal the small white humanoid robot already walking from the road toward the group of men. The robot approaches them calmly from behind. The men remain distracted by their conversation and initially pay little attention to it. The camera operator notices the robot and slightly zooms in, keeping everything imperfectly centered.
3.8-5.5s: The robot walks directly behind one man wearing a black hoodie and dark loose pants. Without warning, the robot quickly crouches down, clearly grabs the back of the man’s waistband with both mechanical hands and gives his trousers one sudden strong downward yank. His dark pants physically slide from the waist down to around his knees, briefly revealing bright colorful boxer shorts. Keep this comedic and non-sexual; no nudity. The man freezes for a fraction of a second in shock, immediately looks down, then turns around toward the robot. His friends notice what happened at slightly different moments and react naturally with surprise, laughter, confusion and angry shouting. Their reactions must feel spontaneous rather than synchronized or staged.
5.5-6.7s: The robot instantly releases the trousers, pivots around and begins running away. IMPORTANT: the prank victim does NOT chase with his pants still down. Before running, he urgently grabs his trousers with both hands and pulls them fully back up to his waist in one fast, believable movement. He adjusts the waistband just enough to keep the trousers secure, then immediately joins the chase. His pants remain properly worn for the rest of the video.
6.7-10.2s: The robot sprints straight across the road toward the camera. All four men now chase after it with their clothing fully in place. The prank victim is visibly angry and runs slightly behind for a moment because he lost time pulling his trousers back up, then accelerates after the robot. The others run naturally behind or beside him.
The men can yell, shout, laugh, gasp or make frustrated vocal reactions during the chase, but the speech should remain indistinct and partially obscured by distance, traffic and movement. No clearly intelligible profanity, no clearly spoken “FUCK,” no explicit slurs, and no clean scripted dialogue. The audio should feel like messy real street reactions rather than readable lines.
The camera operator reacts slightly late, shakily reframing and following the robot. Maintain realistic depth, street perspective, foot contact, clothing physics and human running movement.
The robot runs with fast but physically plausible mechanical steps, arms pumping for balance. Do not make it float or move like animation.
Around 9.8-10.6s: As the robot reaches the near curb, one foot clips or lands awkwardly against the raised curb. It loses balance realistically. Momentum throws its rigid mechanical body sideways and forward. It tumbles onto the asphalt near the curb, rolls once, then lands sprawled on the ground with arms and legs spread. The impact feels heavy and mechanical but causes no gore or visible catastrophic damage.
10.6-12s: The robot makes a small movement as if trying to recover, but before it can stand, the chasing men arrive and surround it.
12-15s: The angry but comedic group crowds around the fallen robot. Several men make quick frustrated kicking or stomping motions toward its hard mechanical body while another bends down and gestures angrily. Keep this clearly slapstick and non-graphic: the target is a machine, no blood, no human injury, no broken bones. The men can continue shouting, but their words remain indistinct and not clearly understandable. The robot remains on the ground while the handheld camera stays several meters away, shaking slightly from the chaos. End abruptly like a genuine viral phone recording.
Overall style: photorealistic real-world smartphone footage, ordinary North American downtown street, storefront windows, concrete sidewalk, asphalt road, parked cars and passing traffic. Natural overcast daylight, realistic clothing and skin, believable crowd reactions, no stylized cinematic lighting. The humor comes from the contrast between an advanced robot and a stupid childish street prank.
IMPORTANT CONTINUITY: exactly one robot; the same four men throughout; the prank victim is the same man from beginning to end; trousers must physically slide downward after the robot pulls the waistband; boxer shorts are only briefly visible after the trousers are pulled down; the victim must immediately pull his trousers fully back up BEFORE chasing; from that point onward his trousers stay securely worn; robot must physically touch and pull the waistband; no teleportation; no duplicated people; no disappearing clothing; no extra limbs; no morphing robot; no superhero movement; no explosions; no weapons; no clear profanity; no clearly intelligible scripted dialogue; no text, subtitles, logos or watermark.
Final Summary
Create a 15-second, 9:16 street robot fight sequence directly from the text prompt. Keep the timing, camera path, dialogue or audio direction, and negative constraints consistent throughout.Want to bring your Street Robot Fight video to life in minutes? Upload your references, paste the prompt, and create your own cinematic AI short with Pollo AI.