Creating AI talking characters with Kling 2.6 Pro - Artlist Blog
Creating talking characters with Kling 2.6 Pro Text to Video  Creating talking characters with Kling 2.6 Pro Text to Video  Creating talking characters with Kling 2.6 Pro Text to Video  Creating talking characters with Kling 2.6 Pro Text to Video  Creating talking characters with Kling 2.6 Pro Text to Video 

Highlights

Learn how to create convincing, cinematic talking characters using Kling 2.6 Pro Text to Video.
Get five ready-to-use prompt examples, explaining why each performance works and how to adapt them for your own stories.
Understand how if you can write dialogue and describe a scene, you can direct believable character-driven videos faster than ever.

Table of contents

Artlist Blog Artlist Blog Artlist Blog Artlist Blog Artlist Blog

Getting characters to talk on screen has been expensive, slow, or complicated. You needed actors, voice recordings, reshoots, or complex animation pipelines. AI video generation changes that.

With Kling 2.6 Pro Text to Video, you can create cinematic characters that move naturally, hold eye contact, and deliver dialogue exactly how you write it. Just a prompt, a voice, and a clear idea of what you want to say.

If you’re a video creator, this opens up real possibilities. You can prototype story beats, build surreal worlds, experiment with tone, or create full narrative moments without waiting on production logistics. Below are five examples you can steal directly. Change the details to fit your project. Swap the setting, the voice, the mood. Let the character tell the story you want to tell.

Example 1: a surreal creature whispering straight to camera

Prompt: A surreal, unsettling light blue shiny plastic horse-like creature with exaggerated sculpted features, long rigid snout, half-closed drooping eyelids with thick lashes, and glossy uncanny-valley toy texture. The scene is dimly lit with a faint spotlight, leaving most of the background in darkness. Slow, almost imperceptible camera push-in. The horse barely moves except for tiny twitches in the eyes and subtle breathing motions. It leans its head forward unnaturally, shadow swallowing half of its face, and whispers in a chilling, soft British accent: “Get this prompt, so you could make me say whatever you want.” The whisper is airy, sinister, almost too close, with subtle reverb. Add slight flicker in the lighting, eerie stillness, and high-detail horror-surreal rendering.

Why it works: Minimal movement, strong lighting direction, and precise vocal tone. Kling handles restrained performances especially well when the mood is doing most of the work.

Example 2: a masked face speaking softly in close-up

Prompt: Close-up of a person wearing a surreal, unsettling crochet mask shaped like a distorted pink flower. The yarn petals are thick and heavy, arranged unevenly, with exaggerated bulges and irregular stitching that feels slightly wrong. The eye openings are small, too tight, ringed with frayed cream yarn, revealing the person’s real eyes staring softly but eerily through the holes. The mask’s round red knitted nose protrudes unnaturally large. The mouth opening is stretched and misshapen, the human lips visible behind it but barely moving.

The lighting is dimmer and colder than studio light, with subtle shadows emphasizing the deep grooves in the yarn. A faint, slow camera drift adds to the uneasiness. The person tilts their head in a slow, unnatural motion and whispers, almost inaudibly: “Get this prompt, so you could make me say whatever you want.” The whisper is soft but eerie, with minimal mouth movement. Hyper-detailed yarn fibers, tactile but unsettling textures. The overall atmosphere is uncanny, quiet, and dreamlike — not cute, not playful, but an eerie handmade creature-mask in a strange liminal space.

  • Style modifiers:
  • uncanny handcrafted wearable mask
  • liminal, eerie, dream-nightmare aesthetic
  • cold desaturated lighting
  • shallow depth of field
  • slow unnatural movements
  • hyperreal knitted textures
  • not a toy, not a doll

Why it works: Close-ups paired with subtle speech keep the illusion intact. You don’t need exaggerated lip movement for impact.

Example 3: a stylish woman walking with poodles and total confidence

Prompt: A glamorous elderly woman walks forward down a narrow outdoor pathway in a pastel-pink neighborhood. She is flanked on both sides by two massive, perfectly groomed bubble-gum-pink Standard Poodles. The dogs have the exact classic show-poodle styling: a large, rounded pom-pom head, huge, rounded side puffs around the ears, a thick, cylindrical body, and tall, fluffy, rounded leg columns. Their fur should be dense, matte, soft, and voluminous, not spherical or stylized.

The woman wears a shiny metallic sky-blue tracksuit consisting of a short, lightweight bomber-style jacket and matching pants. The jacket must NOT be a puffer coat — keep it thin, glossy, and reflective, with subtle fabric wrinkles and pink piping just like a retro satin tracksuit. The pants match the jacket’s exact metallic texture and pastel-blue color. She wears light-colored sneakers, a sky-blue visor, oversized square sunglasses, short blonde hair styled neatly, and she carries a small tan handbag in her right hand.

The environment must replicate a pastel, vintage European-garden residential walkway: tall dark cypress trees lining both sides, evenly spaced; faded pink fences with white railings and ornate metal detailing; a warm, muted color palette with dusty pink walls, light red-and-white painted curbs, and soft golden late-afternoon light. The ground is slightly worn beige concrete, not a bridge, not a boardwalk, not pastel fantasy scenery. Maintain the exact earthy, cinematic, slightly desaturated color tone of the reference.

Camera performs a smooth, gentle forward dolly-in while the woman and the two poodles walk directly toward the lens at the same pace. The framing should match a medium-wide shot: the woman centered, the two poodles walking closely beside her, all three characters filling most of the vertical frame. Keep the perspective natural and realistic—no stylization, no exaggerated features, no oversized hair, no costume changes, no surreal lighting, no pastel fantasy reinterpretation, no fisheye distortion, no environmental drift. Maintain strict accuracy in anatomy, proportions, and materials. Match the reference’s moody, grounded, warm-cinematic look exactly.

The woman says: Get this prompt, so you could make me say whatever you want

Why it works: Natural motion, combined with spoken dialogue, feels instantly believable. This is where Kling excels in lifestyle, fashion, and narrative-driven work.

Example 4: a suited man talking while descending an underground stairwell

Prompt: A man in a suit walks down a tiled underground stairwell at night. The camera is slightly low but NOT extremely low — just below his chin, capturing his face clearly as he descends. His facial features must remain visible, softly lit from above, never turning into a full silhouette. The overhead fluorescent lights glow with bloom, but they do NOT overpower or blow out the image. Exposure stays balanced so the man’s face is partially shadowed but still recognizable, with sculpted highlights on his brow, nose, and cheekbones.

The style must match a soft, dreamy, cinematic thriller: muted green-teal palette, gentle haze, controlled film diffusion, subtle halation, and slightly blurred edges. The lighting is dramatic but not harsh: overhead tubes create directional top-light that shapes the face without hiding it. The tiled walls reflect soft light, adding depth and texture. The image is atmospheric but not dark or horror-like.

Camera movement: a smooth, slow backward tracking shot, always keeping the man’s face in the center frame. The angle stays consistent and never pushes into extreme silhouette. Keep shallow depth of field, soft focus transitions, and high-end cinematography.

Overall mood: elegant, tense, cinematic — matching the exact look of the reference, with visible facial detail, no silhouette, no extreme contrast, and no overexposed light sources

The man says as he goes down the stairs: Get this prompt, so you could make me say whatever you want.

Why it works: Consistent framing and controlled lighting help Kling maintain facial clarity while the character speaks.

Example 5: a woman swinging a bat straight at the camera

Prompt: A gritty low-angle cinematic shot matching the reference style exactly: overhead cool-white ceiling light, muted desaturated colors, dirty basement ceiling texture, soft film grain, wide-angle lens distortion, and a young woman in a pale pink vintage lace-and-ruffle dress, curly messy hair, standing over the camera.

She holds a pink baseball bat with dried blood stains. At the start, the bat is resting firmly on her shoulder, not mid-air, not raised. Her grip is relaxed but ready.

She then performs a full, committed, fast kinetic swing directly toward the camera. The motion must feel heavy and realistic — her shoulders rotate, her torso twists, her arms follow through. The bat accelerates rapidly, passing extremely close to the lens. Use strong motion blur, correct arc physics, and slight camera shake from the force of the swing. The bat should NOT drift, pause, float, or poke; it must behave like a real, heavy bat moving through space with velocity.

Maintain the reference visual style at all times — gritty, low-key lighting from above, muted palette, wide-angle distortion — while ensuring the swing is powerful, smooth, and physically correct

The character says: Get this prompt, so you could make me say whatever you want

Why it works: Action plus dialogue feels bold and confrontational. This is great for trailers, openers, or high-impact storytelling moments.

Make your characters speak for you

Talking AI characters don’t replace filmmakers. They help you test concepts, push style, and tell stories faster than traditional workflows allow.

Kling 2.6 Pro gives you more control, smoother motion, and more convincing speech, all from a text prompt. If you can describe it, you can direct it. If you can write it, you can hear it spoken back to you on screen.

If you’ve been waiting to experiment with character-driven video, now’s the moment.

Start creating with Kling 2.6 Pro on Artlist today.

Artlist BlogArtlist Blog
Was this article helpful?
YesNo

Did you find this article useful?

About the author

Deborah Blank is the Artlist Blog Editor, with over 15 years of experience shaping content for global brands. An expert in AI models, video, and image generation, she’s passionate about empowering creators to tell better stories. Contact her on LinkedIn — she wants to hear from you!
More from Deborah Blank

Recent Posts