Kewin (@Kew_95wtf) header
Kewin (@Kew_95wtf) profile picture
Joined September 2026
Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)Photo by Kewin (@Kew_95wtf)
Tokyo is as tricky as it is beautiful😉👇
2
1
2
66
Image 1 is the outfit, clothes only. Image 2 is the person's face, identity and hair. Image 3 is the app screen; it appears only inside the screen of her phone in shot 5, never floating in the air. CHARACTER LOCK — HIGHEST PRIORITY: the person in every shot is exactly the person in Image 2: same face, bone structure, eye shape, eye colour and spacing, nose, mouth, exact skin tone, freckles and marks, hairline, hair colour, apparent age and build. Do not beautify, slim, de-age or restyle the face. Rested bright eyes, full rested cheeks. Her skin is real and unretouched in every shot: visible pores, fine baby hairs at the hairline, natural texture and a soft natural sheen. She blinks naturally. She is an ordinary woman of average height, always the same size as the other people around her. OUTFIT, the same in every shot: exactly the outfit from Image 1: the cropped green-and-white striped rugby polo shirt with the white collar, the varsity lettering and small patches, showing her midriff; the very baggy faded vintage wide-leg jeans with the burgundy striped tie hanging from the belt loops; the black-and-white high-top sneakers; stacked gold bracelets; and the glossy black leather shoulder bag on her left shoulder. HAIR: exactly as in Image 2, glossy and freshly styled, worn loose, the same in every shot. MAKEUP: polished: soft bronze smoky eyes, defined brows, a warm blush and a glossy berry-red lip. ENERGY: a natural, confident travel creator filming a casual video for her followers: relaxed, warm and charming, genuine smiles, small natural eyebrow raises, easy hand gestures, talking at a lively conversational pace like to a friend. Every expression stays natural and believable. STORY: one day in Japan. In the morning in Tokyo she cannot read any sign; her translation app gets her onto the bullet train to Kyoto; in Kyoto she orders lunch with ease and ends the day at sunset in the old town. A realistic handheld travel film shot on a phone, punchy and fast-paced, quick hard cuts, a different camera angle in every shot, rich natural colours. In every shot one real light source lights her face, named in the shot. Every sign, screen and menu shows Japanese characters. SHOTS, with a hard cut between each: SHOT 1, 0-2 s, TOKYO, MORNING: HIGH ANGLE above the Shibuya scramble crossing in bright daylight. On the first frame she is one small figure standing in the middle of the crossing among hundreds of people of the same size. The crash zoom is already under way on the first frame: the camera zooms in fast and without any pause, past the heads of people crossing in the foreground, and arrives on her waist-up within the first second and a half. The crowd streams past her in SLOW MOTION while she stands still, looking up. Daylight from the open sky. SHOT 2, 2-5 s, TOKYO: LOW ANGLE CLOSE, the camera below her chin looking up past her face to giant video screens and towering signs covered in Japanese characters, their coloured glow on her cheek. She turns her eyes from the signs to the lens, blinks once and says, half-laughing: "Day one in Tokyo. Can't read a thing." SHOT 3, 5-7.5 s, TOKYO STATION: OVER-THE-SHOULDER from just behind her at a row of white train ticket machines under a huge colourful railway route map; each machine has a small touch screen with Japanese buttons, a coin slot and a ticket slot. The touch screen's soft glow lights the side of her face. Her right index finger hovers over the screen, and she says: "Which ticket gets me to Kyoto?" SHOT 4, 7.5-10.5 s, TOKYO STATION: MEDIUM, the camera moving backward in front of her as she walks toward it through a bright, busy station corridor with Japanese signs overhead, the long ceiling lights reflecting on her glossy hair. She looks into the camera and says: "Tokyo is magic. The signs? A total puzzle." SHOT 5, 10.5-16.5 s, SHINKANSEN PLATFORM: CLOSE under cool white platform lights. The camera dollies straight forward at one slow constant speed for the whole shot, at chest height. HER RIGHT HAND HOLDS ONE ORDINARY BLACK SMARTPHONE FOR THE WHOLE SHOT, up close to the lens, the black phone body and its glass screen clearly visible, the screen turned toward the camera and filling the lower half of the frame; the phone screen shows exactly Image 3, sharp, upright and unchanged. Her left index finger points back over her shoulder at the blurred blue platform sign behind her. Above the phone she says: "So I just scan it with this app. Platform seven, to Kyoto. Easy." SHOT 6, 16.5-18.5 s, SAME PLATFORM: LOW ANGLE, the camera at knee height on the platform, in SLOW MOTION. She stands at frame-left behind the yellow platform line while a sleek white bullet train glides in behind her from frame-right, and the rush of air lifts her hair. Cool platform light and the train's bright headlights. SHOT 7, 18.5-21 s, KYOTO, AFTERNOON: MEDIUM at the wooden counter of a small traditional noodle shop, the camera at counter height, wooden menu boards with Japanese characters hanging above her, warm afternoon light coming through a wooden lattice window onto her face. She glances up at the menu boards and says to the camera with a grin: "Spicy miso ramen? Found it." SHOT 8, 21-23 s, KYOTO: EXTREME CLOSE in SLOW MOTION. The chopsticks in her right hand lift glossy noodles out of a steaming bowl of spicy miso ramen with a soft-boiled egg, sliced pork and spring onions resting on the wooden counter, and the steam curls up in the warm window light. SHOT 9, 23-30 s, KYOTO, SUNSET: MEDIUM on a sloping stone-paved street in Kyoto's old Higashiyama district, wooden machiya houses on both sides and the five-storey wooden Yasaka pagoda rising behind her against a golden sunset sky, the warm sunset light on her face. She stands in the middle of the street, her baggy jeans and black-and-white high-top sneakers on the grey stone paving. The street is gently alive: a few locals and tourists stroll up and down the slope behind and around her at a calm distance, two women in colourful kimonos walk past a lantern-lit shop doorway, and a couple stops to take a photo of the pagoda; all of them are ordinary people of normal size, softly out of focus. The camera rises straight up at one constant speed for the whole shot and tilts down to keep her centred, revealing the rooftops of the old town and the people strolling on the street around her. 23-24.5 s: she tilts her face up to follow the lens. 24.5-29 s: she says to the camera with a relaxed smile: "Every sign, every menu, every ticket. Lost in translation? Not anymore." END STATE: a high angle looking down at her on the stone street, her face turned up to the lens, a few people strolling on the stone street around her, the Yasaka pagoda and the old wooden rooftops of Kyoto glowing in the sunset, camera stopped. She has two hands and only two; arms run unbroken from the shoulders. Sound: the crossing chime and a deep, muffled city roar in the slow-motion shot 1; city buzz and screen jingles in shot 2; ticket-machine beeps in shot 3; station announcements and footsteps in shot 4; a quiet platform hum in shot 5; the deep slow-motion whoosh of the arriving bullet train in shot 6; the soft clatter of a small noodle kitchen in shot 7; a gentle bubbling simmer in shot 8; temple bells, birdsong, soft footsteps on stone and quiet voices of passers-by in shot 9. Only her lines in shots 2, 3, 4, 5, 7 and 9 are spoken, each exactly once; nobody else says anything.
2
319
We tried the same storm prompt on Claude Opus 5.5 and Claude Fable 5.2. Every wave, raindrop and lightning bolt is pure code, no 3D models, no footage. Which one would you trust with the boat? Try both on OpenClips
1
1
5
72
Kewin retweeted
POV: Ready for Christmas 🎄🧑‍🎄 Check the prompt in the comments👇
2
2
3
2,360
POV: Ready for Christmas 🎄🧑‍🎄 Check the prompt in the comments👇
2
2
3
2,360
Festive music video, photorealistic, 9:16 vertical, cinematic, one continuous event with cuts. @image1 is the person: keep their face exactly as @image1 in every frame — same face shape, eyes, brows, nose, lips, skin tone, apparent age, build and their own hair from @image1. Preserve their gender presentation exactly as shown. Use they / their / them only. They wear a red velvet Santa look with wide white faux-fur trim, white fur cuffs, a wide black belt with a silver buckle and knee-high white fur-trimmed boots; styled festive look, warm key light on the face. Only this one person sings. 0–2 s: wide shot of a frozen icy vault: above a heavy round steel vault door a red digital countdown display glows, then the door slowly swings open by itself on its hinge, icy fog rolling out. 2–5 s: inside, the person is frozen completely INSIDE a giant clear ice block, eyes closed, ice in front of their face and body. Two pumpkin-headed guards in black suits aim hair dryers at the ice, candle-lit jack-o'-lanterns on the floor; water drips down and cracks spread across the surface. 5–8 s: close-up through the cracking ice: their eyes snap open. They press both palms forward against the ice; the cracked ice breaks apart into big chunks that fall to the floor. They step out, flick their hair over the shoulder, throw their arms wide and start to SING in a powerful, soaring voice, one long held note: "It's tiiiime…" 8–9 s: a bright white flash fills the whole frame. 9–15 s: a cosy snowy Christmas forest set: tall snow-covered pine trees with warm string lights, a big snowman, wrapped presents, snow falling. Six friends in ugly Christmas sweaters, Santa hats and reindeer antlers dance and cheer around them, throwing silver confetti. Medium shot: they hold the final note "…tiiiime!" with one hand raised high in a diva gesture, then laugh and blow a kiss to the camera. Nobody else sings or speaks. Natural skin texture, natural adult proportions, five fingers per hand, natural movement, believable physics: the ice breaks only when their palms push it. Sound: hair dryers humming, ice cracking and shattering, their soaring held note, then sleigh bells, a cheerful crowd and a warm festive whoosh. No on-screen text.
2
79
Kewin retweeted
This would be great way to give away samples of your brand 😂 Check in the comments how I did it👇
1
1
3
31
Kewin retweeted
I don't know if my sense of humor is weird, but I could stopped laughing🤣 Wan 3.0 did pretty good job
1
3
87
Hot take: the backflip is still the hardest test for AI video 🤸‍♀️ Seedance 2.5, Kling 4.0, Wan 3.0 and Kling 3.0 Turbo got one gymnast and one prompt. Keep your eyes on the bottom right and tell us if you saw the flip.
4
2
14
311
I don't know if my sense of humor is weird, but I could stopped laughing🤣 Wan 3.0 did pretty good job
1
3
87
A dachshund on a skateboard just robbed a guy of his juice box, and four AI models filmed the whole getaway 🧃🛹 Gemini Omni 1.2 Flash, Kling 3.0 Turbo, HappyHorse 1.1 and Grok Imagine 1.5 got the same prompt. Which chase would you believe if it showed up on your feed?
1
2
4
299
💚🩷✨😽💫😻
67
260
38
1,976
109,884
Meshy just gave 3D characters bones 🦴 and 500+ moves to use them! Rigging turns a humanoid 3D model into a rigged one. Multi Animation drops up to 10 moves on it in one request. Hip hop. Breakdance. Even a backflip. So we made it a story 💃 Dad has 3 days to learn to dance before the wedding. His 3D twin teaches him every night. Then the slow dance hits a record scratch 👀 Every shot made on OpenClips, straight from Claude through our MCP. First 30 credits are our wedding gift 🎁
4
1
11
345
Kewin retweeted
When you get unlimited you go crazy 😅
1
1
3
29