
Photo by Thought Catalog on Unsplash
The Guide to Prompting Kling 3.0: Tips, Techniques & Prompt Templates
A comprehensive, hands-on guide to writing effective prompts for Kuaishou's Kling 3.0 AI video generation model. Includes copyable prompt templates for multi-shot storytelling, dialogue, camera work, audio design, and more.
Kling 3.0 is the latest flagship AI video generation model from Kuaishou. It represents a major leap in multimodal AI filmmaking — capable of generating cinematic-quality video from text, images, and video references with built-in multi-shot sequencing, native audio generation, and rock-solid character consistency. With support for continuous video up to 15 seconds and up to 6 distinct camera cuts per generation, Kling 3.0 brings an "AI Director" workflow to the table, eliminating the need for manual clip stitching.
This guide covers everything you need to know about prompting Kling 3.0 effectively, with ready-to-copy prompt blocks you can use right away.
Key Advantages of Kling 3.0
- Multi-Shot Storytelling: Generate up to 6 distinct camera cuts with varied angles, compositions, and smooth transitions — all within a single prompt. No more tedious manual editing.
- Extended Duration: Continuous video sequences from 3 to 15 seconds, enabling richer narrative arcs and more complex scene compositions.
- Character & Subject Consistency: Maintain consistent characters, clothing, and visual elements across multiple shots using text, image, or video references.
- Native Audio Integration: Character dialogue, sound effects, and ambient audio with accurate lip-sync generated directly during video creation — supporting English, Chinese, Japanese, Korean, Spanish, and more.
- Multimodal Input: Process text, image, and video references simultaneously for precise scene control.
- Cinematic Quality: Photorealistic output with advanced physics simulation, improved facial emotions, and up to 4K Ultra HD resolution.
The Prompt Formula
Subject + Motion + Scene (optional) + Camera Language (optional) + Lighting & Atmosphere (optional) + Audio (optional)
Think of your prompt as a mini screenplay. Describe how the shot moves, what the subject does, where it happens, and what it sounds like. The more directorial your prompt, the better Kling 3.0 responds.
Building Blocks Explained
| Element | What to Describe | Example Keywords |
|---|---|---|
| Subject | The core object or character | "A young woman", "A black cat", "A vintage car" |
| Subject Details | Appearance, clothing, posture | "wearing a red leather jacket", "with silver-streaked hair" |
| Motion | What the subject does | "slowly turns around", "runs through the rain" |
| Scene | Environment and background | "in a neon-lit Tokyo alley", "on a misty mountain trail" |
| Camera | Shot type and movement | "close-up", "dolly push-in", "tracking shot" |
| Lighting | Light quality and mood | "golden hour", "dramatic chiaroscuro", "soft diffused light" |
| Audio | Sound, dialogue, music | "says in a calm voice", "jazz piano in the background" |
Core Principles
1. Think in Shots, Not Keywords
Kling 3.0 responds best to descriptive, cinematic language rather than lists of tags. Write your prompt like a screenplay direction.
❌ Keyword-style (less effective):
beautiful woman, sunset, beach, cinematic, 4K, dramatic lighting✅ Directorial-style (more effective):
A woman with wind-swept auburn hair stands at the water's edge on a golden beach. The low sun casts long copper shadows across the wet sand. She turns slowly toward the camera, her expression shifting from contemplation to a quiet, knowing smile. Medium shot, shallow depth of field, warm color grading.2. Lead with the Camera
Begin your prompt by establishing the camera's behavior. This gives Kling 3.0 a clear starting point for composition.
Slow dolly push-in starting from a wide establishing shot of a rain-soaked city intersection at night. Neon signs reflect off the wet pavement in streaks of pink and blue. A lone figure in a dark overcoat crosses the street under a transparent umbrella. The camera gradually tightens to a medium shot as they reach the opposite sidewalk.3. Anchor Your Subject Early
Define your main character or object clearly at the beginning. Use distinctive visual features that can be tracked consistently across shots.
A tall man in his early 40s with a neatly trimmed salt-and-pepper beard, wearing a charcoal wool overcoat and burgundy scarf. He stands at the edge of a subway platform, holding a worn leather briefcase. He checks his watch, glances down the tunnel, then looks back over his shoulder with subtle unease.4. Describe Temporal Flow
Tell the model how the scene evolves from beginning to end. This creates coherent motion and natural pacing.
The scene begins with stillness — a porcelain teacup sits alone on a wooden table in an empty café. Steam curls upward in slow spirals. Then, a hand enters the frame from the right, gently lifting the cup. The camera follows the cup upward, revealing the face of an elderly woman who takes a slow, deliberate sip, closes her eyes, and exhales softly.Text-to-Video Prompting
Basic Prompt Structure
Start simple and layer complexity. Here's how the same concept scales:
Level 1 — Basic:
A panda reading a book in a café.Level 2 — Detailed:
A giant panda wearing round black-framed glasses sits at a small café table reading a book. A steaming cup of coffee sits beside the book. Morning sunlight streams through a nearby window, casting warm shadows across the table.Level 3 — Cinematic:
Medium shot with background bokeh and ambient warm lighting. A giant panda wearing round black-framed glasses reads a hardcover book at a cozy café table. A steaming latte sits to the side, wisps of steam catching the golden morning light from a nearby window. The panda slowly turns a page, pauses, and adjusts its glasses with one paw. Cinematic color grading with warm amber tones. Indie jazz softly plays in the background.Motion Intensity Control
Kling 3.0 responds to motion intensity cues. Adjust the energy level of your scenes:
Subtle motion (calm, meditative):
A woman sits in a window seat on a rainy afternoon, her fingers tracing idle patterns on the foggy glass. She breathes slowly, her gaze distant. The camera holds a steady medium close-up, barely moving. Only the rain streaking down the window and the gentle rise of her chest suggest motion.Moderate motion (natural, flowing):
A street musician plays an acoustic guitar on a cobblestone sidewalk in the late afternoon. His fingers move fluidly across the strings as pedestrians walk past in soft focus. A tracking shot follows a leaf blown by the wind from the musician's feet across the frame. Natural ambient lighting with warm golden tones.Dynamic motion (energetic, impactful):
First-person perspective racing through a thunderstorm. The camera plunges through dark churning clouds, lightning cracking on both sides. Rapid rotation as the view breaks through the cloud layer, revealing a sprawling city of lights below. The descent accelerates toward the glowing skyline. Wind roar and rolling thunder fill the audio.Multi-Shot Sequencing
One of Kling 3.0's standout features is generating multi-shot sequences within a single prompt. Number your shots clearly and describe each cut.
Shot-Reverse-Shot Dialogue
Shot 1: Medium close-up of a woman in a blue blazer sitting across a table in a dimly lit bar. She leans forward slightly and says: "You had three months to tell me. Three months."
Shot 2: Cut to a close-up of the man across from her. He's wearing a rumpled white shirt, loosened tie. He swallows hard, looks down at his glass, and responds quietly: "I know. I was afraid of what you'd say."
Shot 3: Cut back to the woman. Her expression softens almost imperceptibly. She sits back in her chair, looks away toward the window, and says after a pause: "That's the problem. You're always afraid."
Shot 4: Wide two-shot of both at the table. A long silence. The bartender moves in the blurred background. Neither speaks.Action Sequence with Cuts
Shot 1: Wide establishing shot of an abandoned warehouse at dusk. Broken windows, overgrown weeds. A figure in a dark hoodie approaches the entrance cautiously.
Shot 2: Cut to interior — low-angle shot looking up at the figure as they push open the heavy metal door. Dust particles float in a shaft of fading light. Their footsteps echo on concrete.
Shot 3: Close-up of the figure's hand pulling back the hood, revealing a young woman with a determined expression and a small scar above her right eyebrow. She scans the space, eyes narrowing.
Shot 4: Over-the-shoulder shot from behind her, revealing rows of old machinery in the dim warehouse interior. A faint red light blinks in the far corner.
Shot 5: Quick cut to the blinking red light — extreme close-up of a small device attached to a support beam. A timer reads 4:32 and counting down.
Shot 6: Medium shot of the woman — she spots the light, and her expression shifts from caution to urgent alarm. She breaks into a sprint toward the device.Narrative Scene with Emotional Arc
Shot 1: Close-up of a child's hand placing a small paper boat at the edge of a stream in a sunlit forest. The boat wobbles, then the current catches it.
Shot 2: Tracking shot following the paper boat as it navigates between smooth stones, spinning gently. Dappled sunlight plays across the water surface.
Shot 3: Medium shot of the child — a girl of about 7, wearing a yellow rain coat — watching the boat drift away. She waves at it, smiling. Behind her, a man (her father) watches from a few steps back, his expression bittersweet.
Shot 4: The boat approaches a small waterfall. Close-up as it tips over the edge, tumbles, and emerges at the bottom, still floating. The girl, visible in the background, lets out a delighted laugh.Image-to-Video Prompting
When using an image as a starting reference, focus your prompt on how the scene evolves from the static image.
Key Principles
- Don't re-describe what's already in the image — focus on motion, changes, and evolution
- Keep movements simple and physically plausible — complex motions from a still image can produce artifacts
- Describe camera movement to add cinematic dynamism even with minimal subject motion
Portrait animation:
The woman in the image slowly turns her head from a three-quarter view to face the camera directly. A gentle breeze lifts a few strands of her hair. Her lips part slightly as if about to speak. The camera holds steady in a medium close-up. Soft ambient light remains consistent with the original image.Landscape animation:
The scene from the image comes alive — clouds drift slowly from left to right, casting moving shadows across the mountain valley below. A flock of birds emerges from the treeline in the distance and flies across the frame. The camera performs a very slow, barely perceptible push-in toward the distant peak.Product showcase:
Starting from the product as photographed, the camera performs a smooth 180-degree orbit around the sneaker, maintaining a consistent medium close-up. The shoe rotates on an invisible platform. Studio lighting creates traveling highlights across the textured surface. The laces sway gently with the rotation. Clean white background, no distractions.Audio Generation Guide
Kling 3.0 generates synchronized audio natively. Here's how to control each element.
Character Dialogue
For accurate lip-sync and voice matching, describe each character's voice distinctively:
Voice description formula: Gender + Age Range + Voice Quality + Speech Rate + Emotional Tone + Language
A woman in her late 20s stands in a sunlit kitchen. Her voice is warm and mid-range with a slightly raspy quality, speech rate moderate, tone playful and affectionate. She looks at someone off-camera and says in English: "I told you not to eat the last piece, and yet here we are."Multi-character dialogue with voice differentiation:
Two colleagues stand at a whiteboard in a modern office. The first speaker is a man in his 30s — his voice is deep and measured, speech rate slow, tone analytical. He points at the whiteboard and says: "If we shift the timeline by two weeks, we can absorb the delay without impacting the launch."
The second speaker is a woman in her late 20s — her voice is higher-pitched but confident, speech rate fast, tone slightly impatient. She crosses her arms and responds: "Two weeks puts us right in the middle of the holiday freeze. That's not going to fly with ops."Multi-Language Dialogue
Kling 3.0 supports precise lip-sync across multiple languages and accents:
Bilingual conversation:
A café in Paris. A French waiter approaches a table where an American tourist sits with a menu. The waiter speaks in French with a polite, formal tone: "Bonjour, avez-vous choisi?" The tourist looks up and responds in accented English: "Um, sorry — do you have anything without dairy? I'm lactose intolerant." The waiter nods and switches to English with a French accent: "Of course. I would recommend the grilled sea bass."Chinese dialect example:
Nighttime at a food stall in Chengdu. Two friends sit on plastic stools eating hot pot. The man on the left, wearing a white T-shirt, speaks in Sichuan dialect: "这个锅底太巴适了,辣得刚刚好。" The woman across from him laughs and replies in Sichuan dialect: "你还说嘛,我都辣得流眼泪了。" She fans her mouth dramatically.Sound Effects
Kling 3.0 generates contextual sound effects automatically. You can also guide them explicitly:
A blacksmith hammers a glowing orange blade on an anvil in a dim stone forge. Each hammer strike sends out a shower of bright sparks and a sharp metallic ring. Between strikes, the low roar of the forge fire and the hiss of heated metal fill the space. The blacksmith pauses, lifts the blade to inspect it, and plunges it into a water trough — a violent burst of steam and a loud sizzling sound.A thunderstorm over an open ocean. Waves crash against the hull of a wooden sailing ship. Lightning illuminates the scene in brief flashes, followed by deep rolling thunder. The wind howls through the rigging. A crew member shouts orders, barely audible over the storm.Background Music (BGM)
Control the mood and style of generated background music through your prompt:
Specify genre and instruments:
A time-lapse of a city waking up at dawn — empty streets gradually filling with commuters, lights turning on in office buildings, coffee shop shutters rolling up. Background music: lo-fi hip-hop beat with mellow piano chords and a soft vinyl crackle texture. Relaxed, contemplative atmosphere.Match music to emotional arc:
An old man walks alone through an autumn park, leaves falling around him. He sits on a bench and opens a locket containing a faded photograph. Background music: solo cello playing a slow, melancholic melody that builds gently as he opens the locket, then fades to near-silence as he closes his eyes and holds the locket to his chest. Intimate, deeply personal mood.Energetic music direction:
A montage of a boxer training — jump rope, heavy bag, shadow boxing, road running at dawn. Fast cuts synchronized to the beat. Background music: driving electronic beat with rising intensity, deep bass drops at each cut transition, building to a crescendo as the boxer steps into the ring under the lights.Camera Work & Cinematography
Camera Angles
High angle / Bird's eye view:
Bird's eye view directly above a circular hedge maze in an English garden. A single person navigates the maze, their progress visible from above as they make wrong turns and double back. The camera holds perfectly still, framing the geometric pattern of the maze. Late afternoon shadows create strong directional contrast.Low angle / Worm's eye view:
Low-angle shot looking up at a towering skyscraper from street level. The building's glass facade reflects drifting clouds and passing aircraft. A pigeon flies overhead in slow motion, crossing the frame diagonally. The sense of scale is overwhelming — the camera emphasizes the building's vertical dominance.Dutch angle (canted):
A detective walks down a long, empty hospital corridor at night. The camera is tilted at a 15-degree Dutch angle, creating unease. Flickering fluorescent lights create a strobe effect. The detective's footsteps echo. The camera slowly dollies backward, keeping him centered in frame as he advances.Camera Movement
Dolly push-in:
Starting from a wide shot of a concert hall stage with a grand piano at center. Warm stage light illuminates the pianist as she begins to play. The camera slowly and steadily pushes in along the center aisle, passing empty seats, gradually tightening from full-stage to a medium shot of the pianist. The movement is smooth and continuous, ending at her profile as her fingers dance across the keys.Tracking shot:
A tracking shot following a cyclist through a narrow European village street. The camera moves at the cyclist's pace at handlebar height. Stone buildings, flower boxes, and hanging laundry pass by on both sides. The cobblestone surface creates subtle camera vibration. An elderly woman sitting in a doorway watches the cyclist pass. The shot continues unbroken for the full duration.360-degree orbit:
A slow 360-degree orbit around a martial artist performing tai chi at dawn on a hill overlooking a misty valley. The camera circles at a consistent medium shot distance, maintaining smooth velocity. As the camera moves, the sunrise shifts from backlighting to side lighting to front lighting, revealing different aspects of the scene. The martial artist's movements are flowing and continuous.Crane shot (rise):
Starting from a close-up of a pair of well-worn hiking boots standing at the base of a mountain trail. The camera slowly rises straight up, revealing the hiker's full body, then the winding trail behind them, then the valley far below, and finally a panoramic view of snow-capped peaks stretching to the horizon. The rise is smooth and continuous, suggesting both achievement and the journey's scale.Shot Scales
Extreme close-up:
Extreme close-up of a human eye. The iris is deep green with flecks of amber. A reflection of a city skyline is visible in the pupil. The eye blinks slowly once, and when it reopens, the reflected skyline has changed — now showing the same buildings in ruins. A single tear forms at the corner of the eye. Macro lens, razor-thin depth of field.Medium shot:
Medium shot of a street food vendor preparing crepes on a circular griddle in a Parisian market. She spreads the batter in a practiced circular motion, adds Nutella and sliced bananas with quick precision, folds the crepe, and hands it to a waiting customer with a warm smile. Natural daylight, shallow depth of field on the vendor's hands and the crepe.Wide establishing shot:
Wide establishing shot of a small fishing village at dawn. Colorful boats line the harbor, their reflections rippling in the calm water. A thin fog layer hugs the shoreline. A few early-rising fishermen prepare their nets on the dock. The camera holds steady, letting the scene breathe. Distant seagull cries and gentle waves lapping against the boats.Visual Style References
Kling 3.0 responds well to explicit style references. Use them to set the aesthetic direction:
Film noir:
Film noir style. A private detective sits alone in his dimly lit office, silhouetted against venetian blinds casting striped shadows across the wall. A half-empty bottle of whiskey sits on the desk. He lights a cigarette, the match illuminating his face for a brief moment — weary eyes, unshaven jaw. A ceiling fan turns slowly overhead. High contrast black and white, deep shadows, 1940s atmosphere.Wes Anderson:
In the distinctive style of Wes Anderson — perfectly symmetrical composition. A hotel lobby in pastel pink and mint green. A concierge in a burgundy uniform stands at exact center behind an ornate front desk. Two matching bell carts are positioned equidistant on either side. The camera is locked on a centered dolly track, pushing in slowly. Flat, evenly distributed lighting. Whimsical yet precise.Studio Ghibli / Miyazaki anime:
In the style of Studio Ghibli. A young girl runs through a field of tall wildflowers on a breezy summer afternoon. Her straw hat flies off and she chases it, laughing. The wind moves the entire flower field in rolling waves. Distant hills and cumulus clouds frame the horizon. Hand-painted texture, soft pastel colors, gentle and dreamlike atmosphere. A whimsical orchestral melody plays softly.Cyberpunk:
Cyberpunk aesthetic. A rain-drenched megacity street at night. Holographic advertisements flicker on every surface — kanji, neon, corporate logos. A delivery drone zips overhead at low altitude. A lone figure in an LED-trimmed jacket walks through the rain, reflected perfectly in the wet road. Steam rises from subway grates. The camera tracks them from a low angle, keeping them small against the towering buildings. Synth-heavy electronic ambient soundtrack.Special Effects & Transformations
Object Transformation
A glass marble sits on a wooden desk in a quiet study. Warm lamplight reflects off its surface. Cracks begin to form across the marble — glowing golden light seeps through the fractures. The marble splits open slowly, and a miniature tree grows from within it at time-lapse speed. Tiny leaves unfurl, small fireflies emerge and orbit the tree. Within seconds, the desk is illuminated by the tree's soft bioluminescent glow. The room's shadows shift as this new light source comes alive.Character Transformation
A woman stands in a moonlit courtyard, her breath visible in the cold air. Her eyes flash silver. Starting from her fingertips, frost crystals begin spreading across her skin like rapid-growing vines. Her dark dress transforms into flowing ice-blue robes that seem to be woven from frozen mist. Her hair lifts and turns white from the tips upward. Ice formations branch outward from her feet, covering the stone ground in intricate crystalline patterns. Her transformation is elegant and controlled — not violent but inevitable, like winter arriving.Environmental Transformation
A barren, cracked desert landscape under a harsh midday sun. The camera holds a wide shot. Then — a single green shoot pushes through the cracked earth at center frame. From that point, verdant grass spreads outward in a rapid radial wave. Flowers bloom, then saplings rise and grow into full trees within seconds. A river appears, cutting through the new forest. Animals emerge — deer stepping out from behind trees, birds bursting from the canopy. The sky shifts from harsh white to a warm sunset gradient. The entire transformation from desolate to paradise unfolds across the full duration.Negative Prompting
Use negative prompts to suppress common artifacts and unwanted elements:
Negative prompt: blurry, low quality, watermark, text overlay, distorted hands, extra fingers, duplicate limbs, unnatural skin texture, overly saturated colors, lens flare, floating objects, inconsistent shadowsCommon negative prompt elements:
blurry/out of focus— prevents soft or unclear renderingdistorted hands/extra fingers— reduces limb artifactstext/watermark— suppresses unwanted overlaysjittery/flickering— helps maintain smooth motionmorphing face— prevents facial distortion during movement
Pro Tips for Better Results
- Think like a director, not a designer: Write prompts as if you're directing a scene on set. Describe what the camera sees, how it moves, and what the audience should feel.
- One main action per shot: Keep each shot focused on a single primary action. Add subtle background motion (breathing, wind, ambient movement) for realism.
- Anchor characters visually: Give characters distinctive, visible features (clothing color, accessories, hairstyle) that the model can track across cuts.
- Number your shots explicitly: For multi-shot sequences, use clear labels like "Shot 1:", "Shot 2:" to help the model parse transitions.
- Specify audio per character: When including dialogue, always indicate who is speaking and describe their voice quality so the model can match lip-sync correctly.
- Use time-based progressions: Phrases like "begins with..., then gradually..., finally..." help the model understand pacing and temporal flow.
- Reference real filmmakers and styles: Mentioning "Kubrick one-point perspective" or "Terrence Malick golden hour" gives the model strong aesthetic anchors.
- Control motion intensity with adverbs: Words like "slowly", "rapidly", "gently", and "explosively" directly influence the energy of the generated motion.
- Keep image-to-video prompts evolution-focused: When starting from an image, describe what changes — not what's already there.
- Iterate relentlessly: Generate, review, and refine. Small adjustments to word choice, ordering, and emphasis can produce dramatically different results.
Conclusion
Kling 3.0 marks a significant milestone in AI video generation. Its combination of multi-shot storytelling, native audio integration, character consistency, and cinematic-quality output makes it one of the most capable video generation models available today.
The key to unlocking its full potential lies in writing prompts like a filmmaker — structured, descriptive, temporally aware, and cinematically precise. Whether you're crafting short films, product showcases, creative content, or visual prototypes, the prompt templates in this guide give you a solid foundation to start producing professional-quality results.
Copy the templates, experiment with styles, adjust for your specific needs, and let your creative vision drive the output.
Author

Categories
More Posts

Top 10 Best AI Video Generators in 2026
We personally tested the top 10 AI video generators in 2026 using the same prompt. Here's how Runway, Kling AI, OpenAI Sora, Google Veo 3, Synthesia, HeyGen, Pika, Luma, Adobe Firefly, and Manus actually performed.


Why OpenAI Shut Down Sora: The Real Reasons Behind the Sudden Exit
OpenAI abruptly shut down its viral AI video app Sora in March 2026, ending a $1 billion Disney deal and raising questions about the future of AI video generation. Here are the three real reasons why.


The Guide to Prompting Google Veo 3.1: Tips, Techniques & Prompts
A comprehensive guide to Google Veo 3.1: prompting tips, realistic skin techniques, no-subtitle tricks, pricing/cost breakdown, and the Veo 3.1 length limit explained. Includes copyable prompt templates.

Waitlist
Early Access
Be the first to know when AcceptPrompt launches. Sign up to get early access and exclusive updates.
Be the first to join. Free early access, 50% off when subscribe. No spam, ever.