AI Prompt KingGet the app
← Prompt library

Airport Action AI Video Prompt: Use Your Photo in Google Flow

Copy the Airport Action AI video prompt to recreate a cinematic barrier vault with your photo. Get Google Flow steps, reference tips and motion fixes.

AI Prompt King · Updated 14 Oct 2026

VideoPrompt + guide9:16 preview

Reference preview; this is not a personalized generated result.

Read the full prompt

AIRPORT ACTION — PHOTO-TO-CHARACTER VIDEO PROMPT Create a 10-second vertical 9:16 photorealistic action clip using the attached action reference and one identity photo. Recast the single foreground performer with the adult person in IDENTITY_PHOTO while keeping the airport setting, obstacle layout and recognizable vaulting action from ACTION_REFERENCE. REFERENCE ASSIGNMENT ACTION_REFERENCE supplies the movement order, camera direction, environment, lighting and staging. Use only the clean action excerpt, not the tutorial presenter, phone recording or website demonstration. IDENTITY_PHOTO exclusively defines the foreground performer's identity: facial structure, apparent age, skin tone, hairstyle and facial hair if visible. Preserve these traits naturally as the head turns. Do not retain or blend in the original performer's face. Do not make a second person from the identity photo. These labels are placeholders. Bind the actual uploaded video and photo with the asset selectors or reference tags available in the interface. CHARACTER AND COSTUME Exactly one adult performer. Keep a complete, anatomically coherent human body with natural proportions compatible with the supplied photo. If the photo only shows the face, infer the unseen body conservatively; do not assume it establishes an exact physique. Dress the performer in a plain off-white sleeveless vest and an ivory traditional dhoti with a narrow gold border, matching the action reference's costume. Keep the clothing consistent throughout. Bare feet have believable ground contact. Preserve the identity photo's hairstyle instead of automatically copying the original actor's hair. Keep fabric folds and loose cloth moving naturally with the jumps. The performer holds exactly one short wooden baton with small gold-colored details. It is a prop, not a broom or a long pole. Keep its length, shape and grip consistent; it does not strike anyone. AIRPORT SETTING A quiet private aviation apron in warm late-afternoon sunlight. Beige hangar walls, low cream-colored concrete traffic barriers, dark asphalt, a parked dark luxury sedan and a stationary white business jet. Keep the barriers and background objects in fixed positions. Leave a clear landing area. Do not add a crowd, airport staff, a moving aircraft, a runway chase or an unrelated fight. COMPOSITION AND CAMERA One continuous-looking side-on tracking shot at roughly waist height, moving along with the performer as they travel past the barriers. Maintain the source's screen direction. Show the performer from head to feet with enough space for the full vaulting arc, planted hand and landing feet. Keep the nearby barrier in the foreground and the hangar, car and aircraft farther behind. Use natural perspective and restrained cinematic motion blur. No selfie view, aerial view, sudden zoom, close-up insert, split screen or tutorial overlay. ACTION PLAN Use ACTION_REFERENCE for the exact gesture order. Adapt its clean two-vault excerpt to the selected 10-second duration without an abrupt speed change. The following phases are guidance, not separate shots: 0.0–1.2 seconds: The performer is already moving toward the first low barrier, leaning forward with the baton controlled in one hand. Keep the opening active and show the obstacle clearly. 1.2–3.6 seconds: Plant the free hand firmly on the barrier, transfer body weight through the supporting arm and swing both legs sideways over it. The torso briefly approaches a horizontal position while clearing the obstacle. Keep the supporting palm attached until the push-off; do not float above the barrier. 3.6–5.0 seconds: Land on the near side with bent knees and believable weight absorption. The cloth settles briefly as the performer takes short steps toward the next barrier. Keep the baton secure. 5.0–8.1 seconds: Perform a second smooth, controlled vault following the reference's arm and leg rhythm. One hand supports the body on the barrier; the legs clear its edge. Maintain correct limb count, joint direction, contact shadows and a continuous body silhouette. 8.1–10.0 seconds: Complete the landing and take a few balanced forward steps. Finish after the landing with the face briefly visible in a natural side or three-quarter view. Do not cut away or end in midair. If the supplied excerpt or supported duration differs, retime these phases proportionally while preserving the reference's order and keeping complete landings. Do not invent an extra vault just to fill time. IDENTITY AND PHYSICAL CONTINUITY Keep the same face through the approach, sideways vault, landing and final head angle. Keep one head, two arms and two legs with realistic joints and stable hands. Avoid face smearing, duplicate limbs, fused fingers, sliding feet, floating fabric, body stretching, a disappearing baton or penetration through the concrete. Preserve consistent body scale and grounded shadows. The car and aircraft remain stationary and retain their shapes. LIGHT, TEXTURE AND AUDIO Warm natural sunlight, soft directional shadows, realistic skin texture, textured concrete, natural fabric and restrained film color. Maintain the source's energetic action-film mood without exaggerated superhero flight. Generate only quiet outdoor ambience, footfalls, light fabric movement and subtle hand-contact sounds. No dialogue, narration, singing, music or imitation of the reference speaker. Do not reproduce the reference soundtrack. OUTPUT One finished vertical clip. No captions, logos, watermarks, phone frame, presenter or website UI. Review the face during movement, both supporting-hand contacts, both landings and baton continuity before downloading.

Recreate an airport barrier-vault scene with your own photo using this Airport Action AI video prompt. The reference shows a single performer in an off-white vest and traditional dhoti moving across low concrete barriers, with a parked car, hangar and private jet behind them. The template separates identity from movement so your photo defines the character while the action clip guides the scene.

Use the Copy prompt button above for the full English prompt. The preview is a short, muted excerpt from the supplied reference, not a personalized output generated with this template. The tutorial narration is newly written for AI Prompt King; it does not copy or clone the reference speaker.

What happens in the reference?

The performer approaches a low barrier, places one free hand on its top, swings the legs sideways across it and lands with bent knees. A short wooden baton remains in the other hand. The movement repeats at another barrier while the side-on camera follows. Warm daylight, cream concrete, dark pavement, hangars, parked cars and a white business jet establish the airport setting.

The original uploaded file also contains a presenter and a phone-screen tutorial. Those sections are not part of the action prompt. Use a clean excerpt containing the action instead of asking the model to reproduce the entire social-media video.

What you need

  • One clear photo of the adult you want to appear in the clip, with a visible face and normal lighting.
  • An action reference you are authorized to use. Choose an excerpt that includes an approach, a complete vault and a landing.
  • A Google Flow video workflow that accepts the required references in your account.
  • The master prompt above, with the reference labels connected to actual uploaded assets.

A front or gentle three-quarter portrait is a useful starting point. A photo showing more of the body gives better appearance guidance than a very tight face crop. Keep only one identity photo initially to avoid conflicting instructions.

Open the clean action reference for an 8.8-second excerpt without the presenter, website recording or original audio. Save it from your browser if you need a local upload.

How to use the prompt in Google Flow

  1. Open Google Flow and create a project.
  2. Select Video mode and look for the Ingredients workflow.
  3. Add your identity photo and a clean action-reference clip if video ingredients are supported by the selected model.
  4. Copy the prompt from this page and paste it into Flow. Assign the photo to IDENTITY_PHOTO and the clip to ACTION_REFERENCE using the actual asset chips or tags shown by the interface.
  5. Choose 9:16. The prompt targets 10 seconds; adapt the action timeline if your selected model offers a different duration.
  6. Check the displayed generation cost and create one output first.
  7. Review the face, supporting hand, legs, baton and landings throughout the clip. Download the result after checking it.

Google's Flow video creation guide describes the Video, Ingredients and Frames workflows. Available controls depend on the model and account. Typing a filename or placeholder label does not attach the corresponding file.

Keep the face and the action as separate references

The identity photo defines who performs the action. The video defines movement, timing, camera behavior and setting. Do not ask the model to copy the original performer's face while also preserving your own face.

The template keeps the reference costume: an off-white vest and ivory dhoti with a narrow gold border. If you want different clothes, replace the entire costume paragraph with one consistent outfit description. Keep the identity photo's hair and facial hair unless you deliberately request a change.

The prop is a short baton. Calling it a broom, a long staff or several different objects in the same prompt can make the model change its shape between frames.

Timing and framing for the two vaults

The prompt uses a 10-second creative timeline: approach, first vault, landing and steps, second vault, then a complete final landing. The clean source excerpt is slightly shorter, so the instruction allows modest retiming instead of claiming frame-perfect motion transfer.

Keep the whole body visible. The shot needs room above the head, beside the swinging legs and below the feet. A tight face crop during the vault hides the movement and makes it harder to judge whether the hands and feet connect naturally with the barrier.

If a short generation cannot handle both vaults, simplify the action to one complete vault and landing. A clear single movement is more useful than a longer sequence with broken anatomy.

What if video Ingredients are unavailable?

Use an image-to-video workflow as an alternative. First create a 9:16 opening image containing your character in the airport setting beside the low barrier. Check the face, clothing, hand position and full-body framing before animating it. Then use that image as the Start frame in a supported Frames workflow.

For this version, remove the ACTION_REFERENCE assignment and describe one approach, one hand-supported lateral vault and one complete landing in text. Keep the identity and setting instructions. A still image does not transfer the original video's choreography, so expect an interpretation of the action rather than an exact motion match.

Common problems and fixes

  • Face changes during the jump: use one clearer identity photo, reduce head rotation and simplify camera movement.
  • Hand floats above the barrier: emphasize a planted supporting palm, weight transfer through the arm and release only at push-off.
  • Legs pass through concrete: reduce the vault height or complexity and keep the entire barrier edge visible.
  • Landing looks weightless: request bent knees, visible foot contact and a short settling movement before the next step.
  • Baton disappears or changes size: specify one short baton in the same hand and reduce unnecessary arm swings.
  • Clothes turn into a different outfit: remove conflicting costume instructions and keep the same reference image.
  • Background moves or duplicates: state that the car, aircraft and barriers remain stationary, and avoid extra camera orbits.

Change one instruction at a time. Prompt wording can guide the output, but it cannot guarantee exact identity or perfect fast-moving anatomy.

Audio and editing for your Reel

The prompt requests subtle ambience and movement sounds only. It explicitly excludes the original speaker's voice and soundtrack. Add your own narration, music you are allowed to use and restrained transition effects in your video editor.

For a tutorial, show a short action excerpt, explain where to find the prompt, demonstrate the references and paste step, then finish with output review and download. Keep the AI Prompt King branding and captions readable without covering the supporting hand or landing feet.

Does copying the prompt generate the video?

No. AI Prompt King supplies the text and guide; generation happens in Google Flow or another compatible tool. Service access, credits and generation limits are separate. Replace the placeholders and attach the actual references before generating.

This is an independently prepared prompt template based on the visible action. A personalized identity-replacement result from this template has not been generated and verified. The reference preview is not proof of a guaranteed result, and no prompt guarantees that a Reel will go viral.

Related prompts and help