灵感图库
探索和使用精选的 AI 图像生成模板 (共 6,519 个)
Surreal Mixed-Media Fashion Portrait Collage
A highly detailed JSON prompt for creating a surreal mixed-media digital collage. It combines high-fashion photography of a crouching model with vibrant, flat vector illustrations of animals (rabbit, deer, wolf) that surrealistically wrap around her body, aiming for a playful, edgy editorial look in a Memphis/Pop Art style.
Golden Hour Noir Car Selfie with Flash
A highly detailed prompt for a high-contrast, glamorous car selfie taken at night with a hard, direct on-camera flash. It specifies the subject's appearance (blonde hair, glass-skin, glossy lips, 90s sunglasses), attire (sequin tube top), accessories (gold jewelry), and technical details (24mm lens, f/2.8, cinematic color grading, hyper-realistic style).
Ethereal Fairy Queen Product Shot
A detailed JSON prompt for generating an ultra-photorealistic, high-end fantasy fashion editorial image. It features a female model posing as a mischievous fairy queen in an enchanted studio setting, delicately holding a glowing, rose-quartz crystal wand. The prompt specifies lighting, wardrobe, background, and camera settings for an ethereal, luminous result.
Hyper-realistic Cobalt Blue Coconut Still Life
A highly detailed JSON prompt for generating an 8K hyper-realistic tropical still life photograph. The focus is a wet, cobalt blue coconut with purple undertones, placed on a rustic wooden stump, using soft natural diffused light and a shallow depth of field (f/1.8).
Y2K Aesthetic Doll-Faced Portrait with Direct Flash
A highly detailed JSON prompt for generating a hyper-realistic portrait of a cute, doll-faced Asian woman in a retro café setting. It specifies Y2K/Korean-inspired makeup, checkered clothing, and uses warm tungsten ambient light combined with a harsh direct flash for a sharp, high-contrast aesthetic.
Elegant Curvy Woman Looking Back Over Shoulder
A JSON prompt focused on generating a photorealistic image of a woman with dramatically curvy proportions, wearing a form-fitting white ribbed dress, standing in an elegant indoor setting at night. The pose involves turning away from the camera and looking back over the shoulder, emphasizing the exaggerated hourglass silhouette.
Ultra-realistic Night Street Scene with Petals
A JSON prompt for generating an ultra-realistic 8K vertical image of an empty street at night. The scene features tall trees lit by streetlights, scattered white petals on the road, and deep shadows, creating a quiet, cinematic ambiance.
Shadow and Glitch Effect Prompt with Color Separation
A structured prompt designed to apply a specific shadow and glitch effect. It requires a reference image for the main subject, which remains clean and centered. Four ghost-like duplicates (Red, Blue, Yellow, Green) are generated behind the subject and are the only elements affected by heavy analog noise, VHS scan lines, and digital glitch artifacts.
Photorealistic Mirror Selfie Portrait with Face Consistency
A detailed JSON prompt for Gemini Nano Banana Pro focusing on generating a photorealistic mirror selfie. The prompt emphasizes high face consistency with a reference image, specific attire (crop top and jeans), detailed makeup, and a bright, softly lit bedroom environment.
AI Transition Workflow: Spinning Dimension Swap
This is a multi-step workflow description for creating a 'Spinning Dimension Swap' effect using Midjourney, Nano Banana Pro, and Kling 01. The Nano Banana Pro step is specifically used to generate a 'Dark Version' of the original shot while maintaining the composition, which is a common technique in advanced AI video workflows.
Ultra-realistic Commercial Fashion Editorial
A highly detailed JSON prompt for generating an ultra-realistic commercial fashion photograph. It features a slender female model with voluminous blonde hair, wearing a white bodysuit and distressed low-waisted denim shorts, set against a minimalist grey studio background with soft, directional lighting, aiming for a high-end editorial quality.
2x2 Grid Scene Progression
A simple text prompt for an image generation model, requesting a 2x2 grid layout that visualizes the chronological events leading up to a specific scene, broken down into 5 hours, 3 hours, 2 hours, and 1 hour prior.
Complex JSON Prompt for Mirror Selfie Scene with Pose and Geometry Locks
An extremely detailed JSON prompt for Nano Banana Pro, designed to generate a mirror selfie scene featuring a synthetic woman in a pink cheeky V-bikini. The prompt uses extensive locking mechanisms for pose, camera angle, and bikini geometry to ensure precise control over the output, allowing only identity replacement.
Hyper-Realistic Selfie Prompt with Technical Camera Specs
A highly detailed JSON prompt designed to generate a hyper-realistic close-up selfie portrait of a young woman in a tropical setting. The prompt includes specific demographic details, attire, pose (winking, tongue out), and extensive technical specifications mimicking an iPhone 17 Pro Max camera, including processing details like Deep Fusion and Smart HDR 6.
Anthropomorphic Fox Character in Winter Scene
A descriptive prompt for generating a cinematic image of an anthropomorphic fox character standing on a snowy bridge at twilight. The prompt emphasizes detailed appearance, warm lighting, and a cozy winter atmosphere.
Illustrated Portrait on Ruled Notebook Background
A detailed JSON prompt for generating an illustrated portrait of a thoughtful young woman sitting cross-legged on a ruled notebook background. The prompt specifies appearance, pose, environment details (including hand-drawn doodles like 'FOCUS!' and 'CREATE!'), and a modern comic-sketch aesthetic.
Soft Grunge Selfie with Black Cat Companion
A detailed JSON prompt for generating a close-up selfie portrait of a young woman in a soft grunge, e-girl aesthetic, holding a black cat cheek-to-cheek. The prompt specifies dark hair, pale skin with freckles, specific makeup (winged eyeliner), jewelry, and warm, moody indoor lighting.
UGC Ad Creative Prompt (TikTok/Reels Style)
A detailed JSON prompt designed to generate a low-fidelity, user-generated content (UGC) style ad creative suitable for TikTok or Reels. It specifies a Gen Z woman with a messy bun, oversized hoodie, and a shocked expression, holding a product close to the camera lens, complete with technical parameters mimicking an iPhone front camera and ring light reflection.
Text-to-Video Prompt: Deadlift with Animated HUD Overlays
A detailed prompt for a text-to-video model (Veo 3.1, used with Nano Banana Pro) describing a cinematic scene of a man performing a deadlift in a gritty gym. The prompt's unique feature is the requirement for neon green Heads-Up Display (HUD) overlays that animate in real-time, displaying metrics like 'Power Output' and 'Form' based on the subject's movement.
Influencer Lifestyle Collage (Two Frames, Same Subject)
A complex JSON prompt designed to generate two distinct frames of the same subject (a young woman in a heather grey lounge set) within a single image, mimicking an influencer lifestyle collage. Panel 1 is a close-up portrait, and Panel 2 is a mirror selfie, with detailed specifications for lighting, pose, and aesthetic (Clean Girl aesthetic, soft focus).
Image-to-Image Transformation: Arm Wrestling with Popeye
A prompt for an image-to-image generation task, instructing the AI to transform an uploaded photo of a person into a high-stakes arm wrestling match. The prompt requires maintaining the subject's face and expression while introducing hyper-realistic Popeye (with massive forearms) as the opponent and Olive Oyl as the referee. This is a creative use of image manipulation via prompt.
Hyper-Real 16K Documentary Shot: The Sun Flickers Off
A highly conceptual and technical prompt requesting a hyper-realistic, 16K documentary-style shot capturing the impossible moment the sun briefly turns off. The prompt demands scientific accuracy in lighting physics, atmospheric scattering, and documentary realism (handheld camera, micro-grain), focusing on the terrifying, instantaneous collapse and violent return of daylight and shadows.
Labeled Diagram of a Human Cell (Comparison Prompt)
A prompt used to test the accuracy and detail of image generation models by requesting a fully labeled diagram of a human cell with a minimum of 10 precise and accurate labels. This prompt is used in a comparison context against another model.
Cinematic Action Scene: Explosion Behind High-Speed Car
A detailed JSON prompt for generating a hyper-realistic, cinematic digital rendering of a high-speed action scene. The scene features a futuristic sports car and three people running away from a massive explosion in a city street lined with skyscrapers. The prompt specifies a low-angle shot, dynamic motion blur, and sharp focus on the subjects against the diffused smoke of the explosion.