Action Scene AI Video Prompt: Bike Jump and Fight
Action scene AI video prompt for Seedance 2.5. You jump a futuristic bike off a ramp, land in fire and fight on a burning highway. 15 seconds, one photo.
Have it running in three steps
- 1
- 2Open Vosu
- 3Paste it and generate
What you'll need
15 min- Vosu
This action scene AI video prompt puts you in the middle of a movie fight. You ride a black futuristic bike toward a broken concrete barrier. The bike flies off it in slow motion and explodes. You land on one knee in the fire.
Then a second woman walks in. You run at her, block her punches and kick her into an overturned truck. The truck blows up behind you, and you stand in front of the fire.
It is 15 seconds, in two parts. You need one clear photo of yourself and one prompt. There is no character sheet this time. The prompt takes only your face, hair and body from the photo and writes everything else in words.
The whole process
- Pick one clear photo of yourself: face, hair and body easy to see.
- Open Vosu, go to video and pick Seedance 2.5.
- Add your photo as the only reference image.
- Set 15 seconds and 9:16.
- Paste the prompt below as it is and generate.
What happens in 15 seconds
| Time | What happens |
|---|---|
| 0–1s | Low camera. She rides the bike at the broken barrier, smoke and fire around her |
| 1–3s | The bike flies off the ramp in slow motion, the burning alien ship high behind it |
| 3–6s | She jumps free, the bike hits the road and explodes. She lands on one knee in the flames |
| 6–11s | The blonde woman comes in. They trade blocks and punches on the highway |
| 11–12s | A high kick to the chest throws the opponent back into the overturned truck |
| 12–15s | The truck explodes. She walks forward with the fire behind her |
Pick your photo
The prompt only takes three things from your photo: your face, your hair and your body type. Clothes, background and pose come from the words. So any clear photo works, even a plain one on a grey wall.
This is the photo we used. It shows the face straight on and the hair from top to end, so the green streaks stay in every second of the video.
Set it up and generate
Open Vosu, go to video and pick Seedance 2.5.
Add your photo as the only reference. Set the length to 15 seconds and the shape to 9:16. Paste the prompt below. It is the exact prompt that made the video above.
[Character] <Rider> corresponds to @Image1. Use only her face, hair and body type. Do not use the background, clothing, pose or any objects from @Image1. The woman in @Image1 is <Rider> and no one else. <Opponent> is not from any reference and must not resemble <Rider>.
[Rider] Clothing: olive-green tank top, dark tactical trousers, cybernetic leg armour with glowing blue LED lines.
[Opponent] A blonde woman with straight shoulder-length hair pulled back, taller and heavier build than <Rider>, black leather jacket over a grey top, dark jeans, black boots.
[Bike] A matte-black futuristic motorcycle: two heavy wide front tires side by side, one rear tire, low-slung chassis, exposed matte metal panels, white LED headlights. <Bike> belongs only to <Rider>.
[Scene] A destroyed six-lane highway filled with drifting smoke, burning car wrecks and an overturned truck, under a heavy overcast sky. The broken triangular hull of an alien ship burns high in the sky far behind the road. Overcast diffused daylight plus orange firelight from the wrecks, volumetric shafts through smoke.
Stage 1 | 0-6 seconds Initial state: the first frame already shows <Rider> riding <Bike> toward a collapsed concrete barrier that forms a ramp, burning wrecks on both sides. Primary event: <Bike> launches off the ramp into a slow-motion mid-air arc, then explodes on impact as <Rider> separates from it and drops into a three-point landing on the asphalt, ringed by flames. End state: at 6 seconds <Rider> is crouched on one knee, right fist on the asphalt, the burning bike wreck behind her in frame left. Camera: single low-angle push-in toward the landing point.
Stage 2 | 6-15 seconds Continue from the previous stage: same rider, same clothing, leg armour still lit, same burning wreck behind her. <Opponent> enters from frame right. Primary event: <Rider> rises and runs at <Opponent> in a rapid, controlled sprint; they meet in close hand-to-hand combat — two clean exchanges of blocks and strikes, ending with a heavy kick to the chest that throws <Opponent> backwards into the overturned truck, which detonates. End state: <Opponent> is down at the base of the burning truck. <Rider> stands in the foreground, back to the fire, holding the final frame. Camera: lateral follow shot, ending on a low-angle medium shot of <Rider>.
[Style] Hyper-realistic live-action look, handheld weight to the camera, real motion blur, practical fire and debris, 35mm lens character. Not a stylised CGI-game look: instead photographic, grounded, physically heavy.
[Audio] <engine howl rising into the ramp>, <metal impact and explosion boom with debris rain>, <body impacts and fabric friction in the fight>, <second heavy detonation with low sub-bass>. (Low, driving percussive score under the whole clip.)
[Maintain Consistency] Keep <Rider>‘s identity, hair, clothing and leg-armour lighting identical across both stages. Exactly two women in the whole video, visibly different from each other, no extra people on the highway. <Bike> appears only in stage 1.
Why this prompt works
Each part of the prompt has its own label: character, rider, opponent, bike, scene. The model reads each one on its own, so the bike does not get mixed up with the truck and the two women do not swap clothes.
The first line says to use only your face, hair and body type from the photo. Your photo can have any clothes and any background. The tank top and the leg armour come from the words.
Each stage has a start, one main event and an end. The end of stage 1 says exactly where she is at 6 seconds: on one knee, fist on the road, the burning bike on the left. Stage 2 starts from that same picture, so the jump and the fight feel like one scene.
The fight is kept small: two exchanges and one kick. Long fights in AI video turn into a mess of arms, and a short one is easy for the model to keep clean.
The style part asks for a photographic, heavy look and rules out a CGI game look. That keeps the fire, the dust and the falls looking like a real film set.
Where it goes wrong
| What you got | The fix |
|---|---|
| The opponent has your face | Keep “<Opponent> is not from any reference and must not resemble <Rider>” |
| You wear the clothes from your photo | Keep “Do not use the background, clothing, pose or any objects from @Image1” |
| A third person shows up | Keep “Exactly two women in the whole video” and “no extra people on the highway” |
| The bike comes back in the fight | Keep “<Bike> appears only in stage 1” |
| The fight turns into a blur of arms | Keep “two clean exchanges of blocks and strikes”. Do not ask for more |
| The leg armour stops glowing after the landing | Keep “leg armour still lit” in stage 2 |
| It looks like a video game | Keep “Not a stylised CGI-game look: instead photographic, grounded, physically heavy” |
| The truck does not explode | Keep “which detonates” and the “second heavy detonation” sound line |
Questions people ask
Which AI model makes this video?
We made it with Seedance 2.5 on Vosu. The prompt is written for it: 15 seconds, two stages, one reference photo.
Can I put myself in it?
Yes. Add your own photo as the reference. The prompt takes your face, hair and body type from it and dresses you in the tank top and leg armour.
Do I need a character sheet?
No. One clear photo is enough here, because the costume is simple and it is described in words. A sheet helps when the costume is complex, like on the fire sorceress page.
Is it vertical?
Yes, 9:16, ready for Reels, TikTok and Shorts.
Is it free?
The prompt is free. Vosu is a paid service with a trial, so the video itself uses credits from your account.
Can I change the opponent or the bike?
Yes. Change the words under [Opponent] or [Bike]. Keep the opponent clearly different from you in hair colour and build, or the two faces may start to look alike.
Ready? Open Vosu, add your photo and paste the prompt.
Why this workflow runs here
Made for short-form
Trial available
One place for the video step
Make yours in the next 15 min
You have the whole prompt and every setting. The only thing left is an account, and that takes a minute.
- Free to start
- Prompts ready to paste
- Upgrade only if you want more
Referral link — if you subscribe I earn a commission, at no extra cost to you. More on that.
Related prompts
Cinematic Free Anime Borscht — The Whole Recipe in One 25-Second Video
One prompt cooks Ukrainian red borscht from raw meat to a served bowl, in glossy anime style. Six stages, hands and food only, vertical for Reels.
Cinematic Free Butterfly Eyes Prompt: The Eye Transition AI Video
A free Seedance 2.5 prompt. A black butterfly flies at the camera, the lights go out, and its wing spots become your glowing eyes. 15 seconds, two reference images.
Cinematic Free Car Jump AI Video Prompt: Your Car Flies Behind You
A car jump AI video prompt for Seedance 2.5. You sit calm in the desert while your own car flies over you and lands in a wall of dust. 12 seconds, 9:16.