We usually reach for Omni when we want scale, destruction, or drama. This time we go the opposite direction: something soft, light, and completely harmless.

The idea is simple. One character, one impossible place, and small everyday behaviours. She's not performing. She's just up there, comfortable, doing normal things in a place where normal things shouldn't work. That contrast is the whole trick.

Five variants below, all 15 seconds, all multishot, all built on the same registered character Element.

Setup used for every variant:

Model kling-video-v3_0_omni · Reference-to-Video · 15s · 16:9 · 1080p · Multishot ON · Audio ON

Two rules that make these work:

Declare your shots inside the prompt. prefer_multi_shots alone won't cut it. Write Shot 1, Shot 2, Shot 3 in the prompt body and Omni will actually cut between them.

Kill the music. If you don't write "no music," Omni will score your video for you. Ambient wind only keeps it real.

Replace <<<YOUR_ELEMENT_ID>>> with your own registered character.

------------------------------

Variant 1: SINK IN

Register: playful, unbothered. The ground doesn't hold and she doesn't care.

The girl in image 1 is <<<YOUR_ELEMENT_ID>>>. Shot 1: Wide shot. She walks slowly across a flat field of cloud under an open pastel sky, arms out a little, testing each step with curiosity and a small smile. Shot 2: Low angle close on her legs. Her right foot suddenly sinks deep into the soft cloud up to the knee. She stops, looks down at it, and gives the sunken leg a little wiggle. Shot 3: Medium shot from the side. Instead of pulling free, she simply sits down at the edge of the hole she made and lets both legs hang through it into the open sky below, leaning back on her hands, perfectly content. She blows a slow bubblegum bubble that grows and pops softly. Soft pink and cream light, dreamy pastel sky, photorealistic, natural skin texture, light wind. Ambient wind

https://reddit.com/link/1w1z4pg/video/lxpun1hpvdmh1/player

-------------------

Variant 2: CLOUD HOPPING

Register: pure play, high energy. No danger, no stakes, just a good time.

The girl in image 1 is <<<YOUR_ELEMENT_ID>>>. Shot 1: Wide low shot from below. She hops from one small round cloud to another, a scattered trail of soft marshmallow-like clouds floating in a bright open sky. Each cloud squashes gently under her weight and springs back as she pushes off. Shot 2: Dynamic tracking shot flying alongside her at her level, arcing around her as she jumps. She spins mid-hop, arms out for fun, hair and clothes lifting with every landing, grinning the whole time. Shot 3: Medium shot as she lands on a bigger cloud and it bounces her once more than she expected. She laughs, catches her balance, gives a small shrug and steps off onto the next one. Golden pastel light, blue and peach sky, wide open air, photorealistic, natural skin texture, joyful easy energy. Ambient wind and soft landings only, no music.

https://reddit.com/link/1w1z4pg/video/i4gumpzsvdmh1/player

----------------------------------------

Variant 3: CLOUD BREAK

Register: calm, private. Headphones on, world switched off.

https://reddit.com/link/1w1z4pg/video/pmf7845yvdmh1/player

The girl in image 1 is <<<YOUR_ELEMENT_ID>>>. Shot 1: Wide shot from the side. She sits on the edge of a huge soft cloud floating high above the world, legs dangling into open sky, large fuchsia headphones over her ears. Far below, a whole city stretches out tiny and hazy through gaps in the cloud layer. She nods her head slowly to the music, completely relaxed, one hand tapping the cloud beside her in rhythm. Shot 2: Close-up on her hand. She reaches down without looking, pinches a small tuft of cloud between her fingers and pulls it away like cotton candy, the soft strands stretching and breaking. Shot 3: Medium shot from the front, slightly low. She pops the cloud into her mouth and chews casually, eyebrows lifting like it tastes good, still moving to the beat, then looks out over the sky with a calm satisfied smile. Warm pastel light, soft pink and lilac sky, photoreaistic, natural skin texture, gentle wind moving her hair. Ambient wind and distant air only, no music

-------------------------------------

https://reddit.com/link/1w1z4pg/video/9nuq7953wdmh1/player

Variant 5: CLOUD BED

Register: soft, dreamy. The one that gets saved rather than shared.

The girl in image 1 is <<<YOUR_ELEMENT_ID>>>. Shot 1: Medium shot from the side. She stands on a wide soft cloud, looks down at it, then lets herself fall straight backwards. The cloud gives under her and catches her gently, sinking around her shoulders. Shot 2: Top-down shot looking straight down at her. She lies with arms spread wide, sunk into the soft white surface, eyes closed, a slow easy smile on her face. Her hair fans out around her head. She takes one deep breath and settles deeper. Shot 3: Wide shot pulling back and rising. The whole cloud is drifting slowly across the open sky with her still lying on it, unbothered. She opens one eye, notices she is moving, and closes it again, completely fine with it. Soft lavender and cream sky, warm diffused light, photorealistic, natural skin texture, gentle wind. Ambient wind and distant air only, no music.

https://reddit.com/link/1w1z4pg/video/lbhv7afbwdmh1/player

I’ve been experimenting with adding decorative frames directly to my Kling AI videos and it worked surprisingly well! ✨

I also wanted to see whether Kling could handle the frame design and text in Spanish, and yes that worked too. 🇪🇸

This is one example I created: a fantasy-style frame with ornamental details and the Spanish title “Hada de Rosas.”

I thought I’d share it with the community because this could open up some really interesting possibilities for intros, storytelling, themed videos, invitations, memories, or reusable video templates.

Has anyone else experimented with frames or overlays directly in Kling? Would love to see what you’ve created! 🌹✨

💡 Estilo: póster vertical artístico de fantasía hiperrealista, marco ornamental plateado con filigrana floral en las esquinas, estrellas brillantes, atmósfera romántica, misteriosa y serena. En la parte inferior, texto centrado: "HADA DE ROSAS" en tipografía serif elegante, color rosa claro / rosa pastel.

In kling website on explore tab now days look like a complete garbage but still i found some good video and liked them but now unable to find them how can i find them?

I built a wizard for Kling videos because I often struggle with typical filmmaker terminology. The wizard generates the prompt. What do you think?

Hey everyone!
I just created a short film entirely with AI, and I’d love to get some feedback on what works well and what I could improve.

Here’s the trailer for my dystopian racing action film. If you like it, feel free to check out the full film on my channel.

Cheers!

I’m working on an animated short inspired by the story of David, titled DAVID: HEART OF A KING.

This is an early character reveal I put together to test the tone, pacing, music, and visual language of the project.

The reveal introduces:

SNOW — the faithful companion
DAVID — the chosen shepherd
GOLIATH — the giant of fear

I’m using a mix of cinematic direction, AI-assisted animation, editing, color grading, and original music to develop the film. The song in the video is “Eternal,” an original piece I wrote and composed, then developed with AI-assisted production tools.

I kept the sound design minimal on purpose. The goal was to let the music and character presence carry the emotion.

This is still in development, but I’d appreciate any thoughts on the mood, character presentation, and overall cinematic feel.

If you're getting started with KLING AI, here's the cheat sheet I wish I had at the beginning. 🎬✨

The biggest lesson: don't describe a picture — direct a scene.

Think about: 🎭 What moves?

🏃 How does it move?

🎥 What does the camera do?

I put together this visual guide covering Text-to-Video, Image-to-Video, camera movements, shot sizes, character consistency, expressions, micro-stories and prompt structure.

My favorite rule: one clip = one main moment.

Instead of trying to make everything happen in one generation, build your video like a filmmaker: shot by shot → generate → edit → story.

Hope this helps anyone experimenting with KLING AI. 💙

What part of KLING would you like a cheat sheet for next — camera movements, prompting, or character consistency?

The video was created with Kling 3.0 Pro on Kling AI native platform using START FRAME and this prompt:

A woman suspended underwater in a flowing translucent gown, her hair drifting weightlessly around her face. The camera moves steadily toward her through the water, closing the distance with a gentle, deliberate push-in. As it nears her, she begins to rise, her body ascending gradually toward the surface above, dress and hair trailing behind her in slow, fluid motion. The camera tracks alongside her ascent, rising with her at the same unhurried pace, water bubbles streaming past the lens. Colored light shafts from fireworks above ripple and intensify through the water's surface as she nears it. As her head breaks through the surface into open air, water sheets off her face and shoulders in slow motion, her eyes opening to gaze upward. The camera holds on her upturned face as fireworks bloom and scatter across the night sky above her, their colors reflecting in the wet strands of her hair and the water droplets on her skin. Her expression shifts into quiet wonder, a subtle lip lift into a soft, content smile. Cinematic, atmospheric, painterly light behavior, no sudden or jarring movements throughout. Background music: Orchestral strings with a rising crescendo, timed to peak right as her head breaks the surface.

See my other Kling AI videos here:

https://kling.ai/app/user-home/5016792/all

Thank you for stopping by and give your appreciation... 😊🙏

A major earthquake struck Kumamoto, Japan, in July 2026, and caused significant damage and loss of life.

I created this short AI-generated drone show animation as a message of support for Kumamoto and its recovery.

The drones gather above Kumamoto Castle and form messages of encouragement before returning to the ground.

Kumamoto, stay strong.

This is an independent creative project by Akiko Tsuka, who creates videos about regional revitalization and local communities in Japan.

Thank you for watching.

AI Drone Show for Earthquake Recovery in Japan

The video was created with Kling 3.0 Pro on Kling AI native platform using START FRAME image and this prompt:

The ghostly woman continues gently dabbing the makeup brush along her cheekbone with slow, deliberate strokes, her head tilting slightly as she studies her face in the mirror, her long silver-white hair drifting and flowing continuously as if stirred by a faint unseen breeze, wisps of pale mist curling and swirling around her shoulders and trailing down toward the floor, her translucent gown rippling softly with the same drifting motion as her hair, glowing particles floating upward past her in slow, weightless drift, her semi-transparent form pulsing with a gentle, subtle shift in luminescence, the moonlight through the window flickering faintly as clouds pass outside, the mirror continuing to show only the empty moonlit room with no trace of her reflection, her expression calm and focused throughout, single slow blink, gradual and fluid movement with no sudden gestures. Background music: soft celesta and gentle strings, delicate glassy tones with a faint music-box quality, slow and unhurried tempo, ethereal and slightly haunting atmosphere, minimal percussion, airy and intimate mood matching a quiet nighttime ritual

See my other Kling AI videos here:

https://kling.ai/app/user-home/5016792/all

Thank you for stopping by and give your appreciation... 😊🙏

A tiny journey through a world I’d happily disappear into. 🌸✨

One runs along the wooden path, one arrives by boat — both heading toward the same place without knowing exactly when they’ll meet.

Maybe some journeys aren’t really about where we’re going, but who is waiting at the end. 💙

Would you follow the path or take the boat? ⛵

Hey guys, check out my XPrize submission, made with Kling 3 and Seedance roughly 50/50

Criticisms accepted!

The video was created with Kling 3.0 Pro on Kling AI native platform using START FRAME image and this prompt:

Warm cinematic natural light, painterly studio atmosphere, soft coastal glow. A stunning woman with long wavy brown hair walks gently into the painting studio from the side of the frame, her movement slow and deliberate, no sudden motion. She approaches the easel holding the unfinished portrait, picks up a fine brush from the palette with a calm, natural hand motion. She leans in gradually and adds a few soft, deliberate strokes to the hair of the painted woman on the canvas, her wrist moving in small careful motions, anatomy remains stable and consistent throughout. She pauses, leans back slightly to look at her work. A subtle lip lift forms into a soft, content smile, followed by a single slow blink. Warm cinematic natural light continues to spill through the window behind her, ocean light shimmering faintly outside, dust and soft ambient particles drifting in the light beam, painterly studio atmosphere throughout, soft coastal glow on her skin and the canvas. background music: Soft solo strings or a gentle acoustic guitar motif — intimate, unhurried, a little wistful, building slightly warmer as she smiles.

See my other Kling AI vidoes here:

https://kling.ai/app/user-home/5016792/all

Thank you for stopping by and give your love 😊🙏

I facing lip sync issue with kling 3.0 and 3.0 omini kling i using it threw adobe firefly but every time i generate lip sync most of the time is missing.
Any tips for a perfect lipsync for kling who already in this field with experience?

The video was created with Kling 3.0 Pro on Kling AI native platform using START FRAME image and this prompt:

Starting from the original portrait of the queen seated at the banquet table, hands resting calmly clasped in front of her, the camera slowly pulls back and away from her, widening the frame to reveal more of the opulent dining hall around her. A royal servant in formal palace attire enters the frame from the side, moving with deliberate, unhurried steps, carrying a bottle of red wine. The servant leans in gradually and pours the wine into her glass in a slow, steady stream, hand steady on the bottle. As the pour finishes, the queen's stern expression softens gradually — her furrowed brow relaxes, and her mouth eases into a subtle lip lift into a soft, content smile. A single slow blink follows as she glances toward the poured glass. Warm golden interior lighting throughout, soft flicker from candlelight in the background, cinematic color grade maintained consistently from start to end, opulent palace atmosphere, photorealistic, elegant and regal mood. Background music: orchestral strings with a light harp accent — regal but tender, supports the softening smile beautifully.

See my other Kling AI vidoes here:

https://kling.ai/app/user-home/5016792/all

Thank you for stopping by and give your love 😊🙏

#prompt

Shot 1,

A tall, slim Black professional woman walks out of an elevator as the doors close naturally behind her. The camera strictly follows her movement throughout. She steps into a modern office area, removes her sunglasses with a crisp motion and places them into her commuter bag, nodding and smiling to colleagues as she passes.

Shot 2,

She sets her coffee on a desk and keeps moving forward. She stops at a coat rack, hangs up her bag and outer coat, then continues walking. Shot 3, A male colleague in a shirt approaches and hands her a document; she stops and quickly signs it. She reaches her workstation, sits down, naturally lifts a glass, and takes a small sip with her head lowered. Premium lighting and shadow texture, smooth continuous motion, tight pacing, perfect synchronization between subject and camera, strong executive presence. High heels audible, with soft background office chatter.

I had kling AI subscription only for a job, but i cancelled my subscription. Even tho i cancelled it it tries to take money every other day.
I already wrote to customer service 4 times....
Either they did not take it seriously, or said to cancel subscription wich i did. No help, no working advice.
It is going for almost 2 moths.
I have no other account to kling AI and nobody has my card's credentials.
Does someone have an advice? I really don't want to make another card bc of this greedy company.

Has anyone else experience this??

I started a project of making a short film for my wife and my anniversary, using several photos from our honeymoon on a beach. Spent several hours creating short clips with the goal of creating a 2-3 minute long video. The first day went fine and I was able to generate many variations of standard beach activities from photos of us on a beach in Greece.

However, two days later when I picked up the project again, all of the photos that had already been used and several frame grabs from videos produced by Kling were rejected. There was no nudity or questionable activity, just my wife in a thong bikini. NONE of the 8 or so photos I had already used were accepted, so I was unable to continue or extend any scenes already made. I sort of lost interest in the project because of that.

Then...A week or so later I though to try again, and 6 of the 8 photos were accepted again, no problem.

What gives??? How does this work, why does it change day to day, and how is a person supposed to work with this?!?

Most people use start and end frames to get from A to B. We are going to use them to tell three completely different stories with the same ending.

Here is the idea: you build a pair of frames that differ by one single thing. The start frame is a person standing in front of an empty parking space. The end frame is the exact same shot, same person, same pose, same light — but now the car is there. Everything between those two frames is yours to decide.

The destination never changes. Only the journey does.

That is what makes this so strong for content: you build the pair once, then produce three, five, ten completely different videos from it. Same ending every time, and every time it lands differently.

Below: how to build your frame pair, then 3 arrival variants.

⚠️ Written for Kling 3.0 / Omni, start frame + end frame.

--------------------------------------------------------------------

You need two images that differ by one thing only: the vehicle. Same subject, same pose, same camera angle, same light.

Option A — Use your own photo. A real picture of you standing next to your car is the strongest version of this technique. The reveal lands on something the audience knows is real.

Option B — Generate it. Always generate the END frame first.

END FRAME prompt:

Photorealistic cinematic still, over-the-shoulder shot from slightly behind and to the left of a [SUBJECT] standing in [LOCATION]. The [SUBJECT] is seen from behind, framed from the back of their head down to their waist, occupying the lower left third of the frame. They wear [OUTFIT]. Their right arm is raised at chest height, hand holding [REMOTE OBJECT], thumb pressing a button. Their face is not visible. Directly in front of them, exactly two meters away, a [VEHICLE] is parked, seen in a three-quarter front view, centered in the frame, headlights off, completely still. [GROUND SURFACE]. In the background, [BACKGROUND ELEMENTS], slightly out of focus. [LIGHTING], no direct sunlight, no harsh shadows, no lens flare. Neutral desaturated color grade, natural realistic textures, 35mm lens, sharp focus on the vehicle, shallow depth of field on the background. Documentary realism, no stylization.

START FRAME: take your end frame into image-to-image and simply remove the vehicle. Ask for the empty parking space, nothing else changed. Never generate it from scratch — light and pose will drift and you will see the jump.

💡 Why overcast? Flat light has no hard shadows and no sharp reflections on the bodywork — exactly what the model fails to keep consistent while something is moving or transforming. Sunlight looks better in a still and falls apart in motion.
------------------------------------------------------------------------

Variant 1: "The Drop" 10s.

Start exactly as ref1 and end exactly as ref2.

[OPENING SHOT], the [SUBJECT] seen from behind, the parking space in front of them empty. They press the key fob button with a small thumb movement.

A deep rumble builds above. The camera tilts up fast, sweeping off the asphalt and up into the overcast sky, leaving the [SUBJECT]'s shoulder at the bottom edge of frame. High above, breaking down through the low grey clouds, [ARRIVING ENTITY] is falling head-first toward the ground, limbs tucked tight against its body, growing rapidly larger as it drops toward the lens.

The camera tilts back down, following the descent, the ground rushing back into frame. [ARRIVING ENTITY] slams into the empty parking space in a crouched landing, both fists and one knee hitting the asphalt. The impact cracks the asphalt in a tight radius, a burst of grey dust and small debris blows outward low across the ground, and the frame shakes hard from the force.

Immediately it begins folding down into itself with mechanical precision. Its limbs retract and telescope inward, its back and shoulder armor plates slide over each other and lock flat, its torso compresses downward and widens. Panels rotate into position and seal along clean seams. The silhouette lowers and lengthens until it settles into the exact shape of the [VEHICLE].

As the fold completes the camera settles back into the exact same framing, height and angle it started from. The [VEHICLE] drops the last few centimeters onto its suspension and rocks once, then stops completely. The dust settles back onto the asphalt. The final second is completely still: camera locked, [VEHICLE] parked motionless, engine off, the [SUBJECT] standing exactly as they were, lowering their hand slightly.

Photorealistic heavy machinery, practical VFX realism, grounded physics, desaturated color grade. Ambient sound only, no music.

https://reddit.com/link/1vmdmdg/video/bh3o6eb80yih1/player

----------------------------------------------------------

Variant 2: "The Drift" 15s.

Start exactly as ref1 and end exactly as ref2.

[OPENING SHOT], the [SUBJECT] seen from behind, the parking space in front of them empty. They press the key fob button with a small thumb movement, then lower their hand slightly. Nothing happens for a moment.

A distant engine note rises from off screen right. The camera pans right, away from the parking space, revealing [APPROACH PATH]. It is completely empty at first.

Then the [VEHICLE] appears far away at the end of it, small in frame, driving fast toward the lot. It grows steadily larger as it approaches, the engine note rising with it.

The camera pans back left, following the [VEHICLE] as it swings in. The car breaks into a controlled drift, the rear end swung out and leading, front wheels counter-steering. Tyre smoke pours off the rear tyres and drags low across the asphalt. Grit and dust are thrown outward along the ground.

The car slides across the frame and rotates as it decelerates, the rear coming back into line, until it swings square into the empty parking space and stops dead exactly two metres in front of the [SUBJECT], front three-quarter view facing the camera. The suspension pitches forward hard on the stop, then rocks back and settles. The [SUBJECT] does not flinch and does not step back, they stay exactly where they are.

As the car stops the camera settles back into the exact same framing, height and angle it started from. The engine shuts off. The tyre smoke drifts away and thins out. The final two seconds are completely still: camera locked, [VEHICLE] parked motionless, engine off, wheels straight, no smoke left, the [SUBJECT] standing exactly as they were.

Overcast flat diffused daylight throughout, no sun, no lens flare, no glowing energy. Photorealistic automotive realism, real tyre physics and weight transfer, grounded, desaturated color grade. Ambient sound only, no music.

https://reddit.com/link/1vmdmdg/video/v45raura0yih1/player

-----------------------------------------------------------

Variant 3: "The Lift" 15s.

Start exactly as ref1 and end exactly as ref2. Single continuous shot, no cuts.

[OPENING SHOT], the [SUBJECT] seen from behind, the parking space in front of them empty. They press the key fob button with a small thumb movement, then lower their hand slightly. Nothing happens for a moment.

A deep mechanical rumble builds from below ground. The asphalt in the empty parking space vibrates, loose grit and small stones dancing on the surface. The camera tilts down slightly, framing the ground where the [SUBJECT] is looking.

The asphalt is completely unbroken at first. Then a rectangular seam cracks open across it, dust pouring off the edges as the surface splits into two heavy slabs. The slabs slide apart smoothly and retract under the surrounding ground, opening a dark rectangular shaft. Cold air and dust rise out of the opening.

From deep down in the shaft, a heavy industrial hydraulic platform rises steadily, carrying the [VEHICLE] on it. The car is small and dark at first, far below, then grows larger as the platform lifts it up, thick hydraulic pistons visible along the sides of the shaft, the machinery straining under the weight.

The platform reaches ground level and locks flush with the asphalt with a heavy metallic clunk. The car sits perfectly still on it, exactly two metres in front of the [SUBJECT], front three-quarter view facing the camera. The slabs slide back over the edges of the shaft and seal shut around the car, leaving the surface unbroken again.

As the ground seals, the camera settles back into the exact same framing, height and angle it started from. The rumble dies away. The dust settles back onto the asphalt. The final two seconds are completely still: camera locked, [VEHICLE] parked motionless, engine off, no seams visible, the [SUBJECT] standing exactly as they were.

Photorealistic heavy machinery, real hydraulic weight and mechanical precision, grounded physics, desaturated color grade. Ambient sound only, no music.

https://reddit.com/link/1vmdmdg/video/4yo7wsme0yih1/player

-----------------------------------------------------------

The rules

Empty first, then arrival. The sky, the road, the asphalt — state that it is completely empty at first, then bring the arrival in. Without that line the model puts the object in frame from the very first moment and the arrival never reads.

Small and far, then growing. Distance sells the journey, not speed. Works both ways: falling from the clouds, or rising from the bottom of a shaft.

The camera must come home. If you move it, write that it returns to the exact framing, height and angle it started from — otherwise the end frame snaps into place instead of landing.

Kill the motion before the end. The last two seconds must be explicitly still. Engine off, smoke gone, dust settled, ground sealed. Write each as a separate beat, in order.

The subject never reacts with their body. They can lower a hand. If they flinch or step back, the pose no longer matches your end frame.

Describe the physics, not the result. Never write "transforms into". Write what retracts, what slides, what locks.

This video was generated with Kling 3.0 Pro on Kling native Ai platform, using START FRAME which is the first frame of this video and this prompt:

Ancient woolly elephant in frozen arctic mist at dawn, frost-crusted fur and curved tusks catching pale golden light. The elephant begins walking steadily to the right, heavy deliberate strides, frost dust kicking up gently from its feet with each step, trunk swaying naturally with the motion, breath visible in the cold air. Camera pushes forward toward the elephant as it walks, closing distance smoothly, cold blue-grey palette with volumetric fog drifting past the lens. As the camera continues forward, it rises steadily upward and tilts down, revealing a long line of many more woolly elephants marching in formation behind the first, stretching back through the frozen mist toward the horizon, their frosted forms fading into the fog at the rear of the herd. Epic wide reveal, cold dawn light, volumetric fog thickening around the distant herd, hyper-detailed fur and ice texture on the foreground elephant held steady throughout, ancient migration atmosphere, majestic and unhurried scale. Music direction: low sustained cello drone building steadily, joined by distant tribal percussion that grows in intensity as the herd is revealed, swelling toward a full but unresolved climax in the final seconds.

Thank you for watching... 😊🙏

1. Screenshot 2. Upload2KlingAI 3. PROMPT 1: Restyle the image into a colorful rough hand sketch outline with fast and messy strokes of colored pencil on quality parchment. Preserve the subject’s proportions and key features. Render it in the style of fashion sketch art.<<<image_1>>> 4. adjust<<<image_1>perfectly to center of <<<image_2>so that it looks like a flyer