Catch me if you can

Recreated 9 Sep 2026

Inspired by @frametheory058on XTheir workflow. Rebuilt step by step on OpenClips.

14122.6K

Make it yours

Recreate workflow

Walks the chain step by step — nothing spends until you run one

Built from1 in

Prompt1,110 chars

Cinematic travel-fashion chase film. @avatar is the lead — a real person filmed on location, exact face and identity kept consistent throughout, always wearing white wired earphones. @avatar2 appears at the very end as a friend. Concept: @avatar teases the camera, then runs — and the camera chases. Each time they move, a real in-world event (a passing bus, a splash, an elevator door, a thrown jacket, a tram, a reflection) carries us straight into a brand-new location: rain-wet city, Mediterranean old town, alpine valley above the clouds, golden-hour coast, neon night, misty mountain, blue-hour plaza. The camera loses them in a crowd — then a tap from behind, they're standing right there with @avatar2, and the two run off together as a clean movie-title card reads CATCH ME / IF YOU CAN. Feature-film look: natural handheld motion, 35mm grain, rich color arc from cool rainy city to warm coast to neon and blue hour, realistic skin and movement. Every transition physically motivated — no morphs, no portals. Premium cinematic-electronic track synced to the cuts, layered real ambience, no dialogue.

How it was made

Choose or generate an avatar

Video generation

Photoreal cinematic travel-fashion film. The LEAD is @image1 — preserve @image1's exact face and identity perfectly throughout, real skin texture, real pores, individual hair strands, natural expressions, realistic anatomy; a real person filmed on location, never a beauty-filter or AI look; no garment, colour or accessory description in words. THE FEATURED ITEM is exactly what is shown in @image2 — use @image2 exactly as provided (its real shape, colour, materials, hardware and details; do not describe, rename, recolour or restyle it). @image1 wears or carries @image2 on the body in the natural way that item is worn, the SAME @image2 in every single shot, and it stays clearly visible in most shots. ITEM LOCK: @image2 stays consistent and physically believable at all times — worn/held naturally and securely, moving with real weight, never morphing, changing, duplicating, clipping through the body, or disappearing; whenever @image2 catches the light, let it read clearly without turning into an obvious product shot. The friend who appears at the end is @image3. STORY — CATCH ME IF YOU CAN. CITY HOOK: rain-wet city street, cool cinematic daylight, intimate 24mm close-up; @image1 stands inches from camera, playful knowing smile, @image2 clearly with them, then says out loud in clear English, teasing and close to the mic: "Catch me." — turns and runs, camera chases handheld, @image2 steady and visible; a red city bus crosses extremely close and covers frame (practical wipe). MEDITERRANEAN OLD TOWN: same running stride into an atmospheric Mediterranean old-town market, pale stone alleys, real pedestrians, textured stone, warm shopfronts; @image1 looks back laughing, @image2 still clearly with them, jumps across a puddle, camera dives toward the splash (water fills frame). ALPINE REVEAL: splash match-cuts into the reflective floor of a glass elevator; @image1 enters, @image2 visible in the mirrored reflection, doors close then reopen onto an enormous alpine valley above the clouds; a smile back over the shoulder and @image1 runs outside, cool mist, huge depth, realistic wind. ADRIATIC COAST: match the stride onto a golden-hour coastal road, ocean and cliffside architecture behind, camera tracks alongside; @image1 spins and laughs as the low sun catches @image2; a passing open-top car wipes frame (occlusion transition). NEON NIGHT: frame clears into a rain-soaked neon city at night, deep blacks, cyan reflections, controlled red practical lights, steam, natural crowds; @image1 walks confidently toward camera, neon glinting off @image2; a passing tram blocks frame (occlusion transition). ABOVE THE CLOUDS: tram windows match-cut into a mountain cable-car window; @image1 stands against the glass over mist-covered peaks, reflection overlaying the landscape, @image2 visible, music goes quiet; @image1 notices the camera in the reflection, mischievous smile, exits frame, camera whip-pans after them. WE LOSE THEM: whip resolves into a lively blue-hour plaza, camera urgently searches moving crowds; @image1 appears, disappears behind pedestrians, vanishes, music near silence; someone taps the camera from behind, fast whip around — @image1 stands behind the camera with @image3, both laughing, beat drops and they run away together. HERO ENDING: camera chases @image1 and @image3 across the plaza, they run, spin and laugh naturally, a flock of pigeons lifts ahead; introduce huge premium warm-ivory condensed typography CATCH ME then IF YOU CAN, people naturally overlapping the letters for real depth, no glow or gradients, clean theatrical movie-title design; then a brief, understated final insert — a natural close shot where @image2 catches the light for a beat (subtle, not a hard product shot) — @image1 runs toward camera laughing and covers the lens with a hand; hard cut to black on the final beat. CINEMATIC LOOK: feature-film cinematography, 24mm chase, 35mm movement, 50mm portraits, natural handheld micro-movement, 180-degree shutter, realistic motion blur, fine 35mm grain, subtle halation, soft bloom, expensive highlight roll-off; color arc cool rainy city, warm earthy town, cyan misty alpine, golden coast, cyan-red neon, cool mountain, rich blue-hour finale, natural skin throughout. TRANSITION RULE: every transition physically motivated (bus wipe, splash, elevator doors, car wipe, tram wipe, reflection-whip, crowd reveal, hand-to-lens); never morph locations, no portals, no liquid AI transforms, no speed ramps, no impossible camera moves. REALISM LOCK: same exact @image1 identity throughout, same @image2 throughout, real skin texture, correct hands, natural running, weighted hair and clothing, physically correct reflections and splashes, background people move independently and never freeze or duplicate, no warped architecture or plastic skin. SOUND: @image1's single spoken line "Catch me" at the very start is clear, close-mic and perfectly lip-synced, then no more dialogue; under it an original 120 BPM premium cinematic-electronic travel track tightly synced to every transition — building through the chase, uplifting hook at the coast, deeper at neon night, stripped during the cable-car and crowd search, massive beat drop at the friend reveal, strongest for the final run; layered realistic footsteps, traffic, splash, elevator ding, wind, tram rumble, cable-car hum, crowd ambience, pigeon wings and genuine laughter; no other dialogue, no voiceover, no lyrics. Negative: describing @image2 or any garment/accessory in words instead of following the reference; @image2 changing, morphing, recolouring, duplicating or clipping through the body; any age, gender or ethnic description; warped faces, distorted or extra hands or feet, plastic skin, watermark; muted or silent speaker at the opening line; lip movement without audible voice; morphing or portal transitions; frozen or duplicated background people.

ModelSeedance 2.5

Aspect9:16

Duration30s

Steps2

SourceX

Inspired by @Diplomemeon X

A punchy 15-second UGC-style sneaker ad shot on a phone in bright daylight — your avatar unboxes a fresh pair of sneakers, laces up in activewear, and tests them out walking and jogging around the city before a quick thumbs-up to camera. Fast authentic cuts, real smartphone look, upbeat energy.

2 steps943K

Seedance 2.515s · 9:16

Inspired by @aiwithlumion X

Live-action rustic product film with glowing golden line-drawings rising out of the steam above the mug.

2 steps1999.5K

Seedance 2.510s · 9:16

Inspired by @Arina_hoqeon X

Generate a video of your avatar with 3 different product.

6 steps33718.1K

MiniMax H315s · 9:16

Inspired by @Goodmanprotocolon X

A trendy 16:9 fashion reel where one person styles the same shirt — the exact garment from your reference photo — in 10 completely different ways: open, knotted, tucked, belted, layered and more. Fast whip-pan and snap-cut transitions between looks, bright minimalist interiors, natural confident posing, upbeat music, and a final reveal of all 10 looks together. Keep the same person and the same garment throughout; the garment's colour, pattern and cut come only from the reference.

2 steps

Seedance 2.527s · 16:9

Inspired by @Its_Saad_kkkon X

Generate a 15-second skincare hero film with your product and your avatar.

3 steps20714.6K

Seedance 2.515s · 16:9

Inspired by @1H77kon X

Generate an action commercial video for noodle products.

2 steps2396.7K

Seedance 2.515s · 9:16

Inspired by @Chaemate_on X

Generate a 30 second vlog video through a lantern-lit riverside festival.

2 steps2723.4K

Seedance 2.530s · 16:9

Inspired by @saniaspeaks_on X

Casual smartphone-style video (9:16), filmed handheld by a friend following @avatar (identity locked). @avatar visits a busy open-air market, browses the stalls, chooses the two products from the attached references, chats with the seller, buys them into a clear bag, tries one, then walks home and places the shopping on a kitchen table. Short English lines ("I came to do some shopping today!", "Wow, there is so much here!", "Wow, it is really good!"). Opens on their face turning to the friend's camera, not on their back. The two products come only from the references and stay identical in every shot — same items, same amount, same bag. Natural handheld shake, autofocus, realistic physics; real market ambience + their voice, NO music.

2 steps12013.1K

Seedance 2.530s · 9:16

Inspired by @HeyAbhishekon X

Generate a 15 second video, with yourself, sharing the last bite with a stray animal.

4 steps35633.9K

Seedance 2.015s · 16:9

Inspired by @PixelAigcon X

Seoul street at night, medium close-up. @avatar walks away from the camera along a narrow street lit by defocused neon signage and shop light, wet asphalt throwing reflections. The camera follows from just behind at @avatar's eye level, handheld with subtle natural instability — never jittery. @avatar turns its head back over its shoulder to glance into the lens and smiles, lifts its right hand and brushes its hair back from her face, then turns forward again and continues walking, drifting to the left of frame. Each of these happens exactly once, in that order — no repeated glances, no looped gestures.

2 steps19326.9K

Seedance 2.510s · 9:16