Welcome to the new AI-powered image visual creation experience! This update introduces groundbreaking optimizations for a seamless creative experience — no more switching between multiple tools or features. You can now effortlessly move from basic image generation to advanced detail editing, dramatically enhancing the efficiency of turning ideas into reality. Here, you simply describe your ideas in natural language and easily activate full-scene creative capabilities: you can perfectly depict an image from pure text or upload reference images to achieve precise editing and modifications on a single image. Up to 10 reference images' features can be integrated for re-creation, fully satisfying all your image creation needs in one go.

📌

🎉 Kling launches a new versatile image generation model - Image 3.0🌟

Welcome to the new AI-powered image visual creation experience!

This update introduces groundbreaking optimizations for a seamless creative experience — no more switching between multiple tools or features. You can now effortlessly move from basic image generation to advanced detail editing, dramatically enhancing the efficiency of turning ideas into reality. Here, you simply describe your ideas in natural language and easily activate full-scene creative capabilities: you can perfectly depict an image from pure text or upload reference images to achieve precise editing and modifications on a single image. Up to 10 reference images' features can be integrated for re-creation, fully satisfying all your image creation needs in one go. The model offers four core advantages:

Feature Retention: Ensures that main elements stay stable without deviation.
Precise Detail Modification: Every adjustment is executed as expected.
Accurate Style Control: Ensures a unified atmosphere throughout the image.
Rich Imagination: Brings your creative visions to life with added tension.

This model operates based on the Multi-modal Visual Language (MVL) concept, using a combination of natural language + images as the core interaction interface. With a robust knowledge base and visual reasoning ability, the model can accurately interpret your comprehensive intentions, ensuring both feature consistency and the stunning display of creativity, all while offering an intuitive and efficient experience.🤩👏🏻

Key Features of IMAGE 3.0

1. High Feature Retention

📷

IMAGE 3.0 supports extracting features from up to 10 reference images, precisely locking in the subject contours, core elements, and tonal qualities of each image. Even when generating across images or in series, it maintains consistency—ideal for IP character design, comic scene creation, and brand consistency in visual design, boosting the efficiency of content production.

2. Precise Detail Modification

📷

With IMAGE 3.0, no professional editing skills are required. Simply provide text instructions to add, remove, or modify objects and characters in the image. The model precisely responds to edits while preserving the original style, lighting, and texture, making the result natural and cohesive. This significantly reduces the number of steps compared to traditional editing tools, boosting creativity efficiency.

3. Accurate Control of Style and Tone

📷

IMAGE 3.0 can deeply deconstruct the core style of reference images, precisely capturing the brush strokes, color combinations, and composition logic to achieve seamless restyling and natural blending of styles. Whether it’s felt, 3D figurines, or niche trends, the model can faithfully reproduce and meet various demands, such as brand visual consistency, cinematic scene design, artistic style transfer, and personalized creative expression.

4. Rich Imagination

📷

IMAGE 3.0 facilitates the natural and harmonious blending of reference images, supporting doodle editing and creative idea triggering via natural language commands. It accurately captures the core features of different reference subjects, preventing forced or awkward combinations, and presents seamless integration. The model supports creative concepts, even for complex ideas.

Usage Scenarios

1. Filmmaking

Multiple-angle Views of Characters

Camera angle/height Modifications

Shot Size Modifications

Expression Modifications

Lighting and Shadow Modification

Color Tone Modification

2. Portrait processing & stylization

ID/Portrait photo

Portrait photoshoot

Multiple characters/group portrait

Avatar stylization

3. E-commerce content production

Product display

Virtual try-on

4. Other Creative Uses

Prompt Handbook

1. Prompt Writing Techniques

Image Editing – Prompt Tips

Formula: Keep everything else unchanged, modify [ ] to [ ].

Example: Keep everything else unchanged, change the cat into a dog.

Editing Target: Specify the object to be edited, e.g., subject, background, angle, color, etc.

Change：Describe the desired effect after editing, e.g., delete, add, make it smile, etc.

Face Reference – Prompt Tips

Formula: Keep []’s facial features, character has [] hairstyle, wearing [] clothing, performing [] action, background is [], with [] lighting, and [] atmosphere.

Position Reference - Prompt Tips

2. Vocabulary Table Example – Portrait Editing

Beauty Enhancements

Hairstyle:

Actions:

Outfit:

Element Editing

Add

Remove

Arrow / Selection / Doodle Editing

Change Background

Material Modification

HD Restoration

Reverse Mapping

Other Modifications


Coloring the line art.	Keep the logo unchanged and generate a mug and packaging box with the logo, primarily in black and white.	Change the background to a Ferris wheel, and transform a man sitting on a bench with an umbrella into a bouquet of balloons.

Enlarge the cup in a girl’s hand, making it enormous, with her holding it with both hands.	The female model is wearing brand sunglasses, sitting on a white chair, with the background being image 4.	Fisheye lens shot, with the subjects from image 1 and image 2 taking a selfie together, with the background featuring a New Year's-themed cityscape, and the girl from image 2 crouching.

Reference Image
	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2

	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2	Change the style of Image 1 to that of Image 2


Convert a photo into an overhead view and mark the photographer's position.	Generate a picture from an arrow-pointed perspective with proper lighting and add a cake to the table.	Imagine what the girl will look like 10 years from now.

Convert the anime-style character in the image into a realistic human portrait.	Generate a realistic 3D-style render/model.	Recreate the object based on the design details shown in the image.

Merge the animals from image 1 and image 2.	Light the scene according to the color and direction of the arrow, then add a cake on the table.	Keep the cat's fur color and transform it into a character inspired by Nezha.

Reference Image
	Keep the girl’s facial features consistent with Image 1. She holds a transparent umbrella decorated with snowflake patterns, wearing a thick vintage-patterned coat and a plaid scarf. The background is a black night sky filled with colorful, glowing star-like light dots, creating a dreamy and romantic atmosphere. Film texture, rich dreamy colors, hazy lighting, heavy grain, vintage aesthetic, facial lighting, CCD look, film grain.	A bun hairstyle, silver earrings, and a brown fur coat. Loose jeans, black boots, red knitted gloves, and a brown leather bag. The girl sits on a white bench against a gray brick wall, lifting her head and smiling happily at the camera. Shot from a top-down perspective, with casual and carefree posture.	A bun hairstyle, silver earrings, and a brown fur coat. Loose jeans, black boots, red knitted gloves, and a brown leather bag. The girl sits on a white bench against a gray brick wall, lifting her head and smiling happily at the camera. Shot from a top-down perspective, with casual and carefree posture.	Real-person photoshoot, keeping the character’s facial features unchanged, with various poses and expressions in realistic photography. Long deep-brown hair with loose, voluminous waves. Background: classic Disneyland scenes such as in front of the castle, beside the carousel, or Mickey Avenue. Bright colors, fairytale atmosphere, blurred background to highlight the subject. Tone: vintage film style, warm and soft hues, slight grain. Half-body shot with the subject positioned in the lower-middle of the frame. Wearing a red-and-white polka-dot Minnie headband/hair clip and large plush Mickey gloves.	The girl wears a thick white sweater, a white checkered scarf, red gloves, and a red knitted hat. She holds a sparkler in each hand, standing at the foot of a mountain. Huge fireworks burst in the sky above her. A breeze blows past, moving her hair gently.

	Warm winter atmosphere. The girl has long wavy hair with a black velvet bow on her head. She wears a red sweater, black skirt, black leather shoes, and white lace gloves. She slightly turns her body while crouching inside a giant crystal ball with snowflakes falling inside. The background includes a Christmas tree and small snowman decorations. Christmas atmosphere, realistic style, film-like texture, visible grain.	Long hairstyle, wearing a light-blue silk camisole and white lace shorts, with feathered wings on her back. She wears pink lace-up ballet shoes and lies on a reddish-brown velvet sofa, holding a book and resting her chin on one hand. Playful and relaxed. Vintage style with warm color tones.	A gray trench coat, long woolly curls, holding a bouquet of fresh flowers, sitting on top of a car. Green plants and a utility pole in the background. Korean photography style.	Exquisite makeup with delicate skin texture, fair porcelain-like complexion. She looks into the camera with rosy cheeks and a soft ceramic glow. Her hair is braided and pinned back, decorated with light green ribbon bows. She sits elegantly on a vintage white floral chair, holding a glass vase with pink roses in her right hand. Outside the wooden window frame are lush white, pink, and multicolored rose bushes. Soft light falls on her face, body, and hair, making the strands glow. French vintage style with forest fairytale vibes, dreamy and romantic portrait photography, cinematic film texture.	The girl has long curly hair and faces away from the camera, wearing a backless lace dress. She looks into a vintage vanity mirror, and her face is visible through the reflection as she gets ready. Sunlight shines onto her back, highlighting her smooth skin. Vintage decorations hang above the dressing table. European-style retro bedroom setting.

Reference Image		Reference Image
Merge Image 1 and Image 2 into a single overhead selfie of two people. The composition should be tight, with the two subjects close together. Their heads are slightly tilted upward, eyes looking directly at the camera, creating a strong visual impact. The person on the left stands slightly forward, keeping the character design and features from Image 1 unchanged, maintaining facial similarity. The person on the right follows the character design and features from Image 2, keeping facial similarity, with the body slightly inward. High-angle overhead shot, exaggerating the head proportions, typical of Japanese/Korean selfie style. Background is pure white, simple and clean, emphasizing the subjects. Overall style leans toward Japanese visual style, high image clarity, shot with iPhone front camera, producing a refined, fashionable, slightly stylized group photo effect. Ensure seamless integration of both subjects, natural visual transition, and bright, even lighting across the entire image.		The boy from Image 1 imitates the fox pose from Image 3, and the girl from Image 2 imitates the rabbit pose from Image 3.
Reference Image		Reference Image
Elegant posture, standing next to a railing. One hand wraps around the waist, while the fingers of the other hand point toward the chin. Romantic bridge atmosphere, silhouette effect.		Use a fisheye lens. Create a selfie-style group shot of the subjects from Image 1 and Image 2, with a city background featuring a festive New Year atmosphere.

Kling IMAGE 3.0 Model User Guide