Curated AI prompts for stunning image and video generation. Updated daily from top creators.
This is a prompt used to generate a combat scene, comparing the output quality between GPT-image and Nano Banana Pro, using the same prompt and reference image.
Prompt Preview
I use GPT-image for drawing combat scenes ( ´∀｀) Left: GPT-image Right: nano banana pro Same prompt, same reference image
A template for 'Forbidden Items' (negative prompts) specifically tailored for generating manga panels in NanoBanana Pro, ensuring correct reading order, language, and format.
Prompt Preview
[Forbidden Items] - Left-to-right reading, bottom-to-top reading composition is forbidden - English dialogue is forbidden - Horizontal writing of dialogue is forbidden - Monochrome, black and white, line art is forbidden
A creative JSON prompt for generating a LEGO miniature diorama scene depicting the film set of the horror movie 'The Nun'. It includes detailed descriptions of the gothic stone interior, the LEGO nun character, and various LEGO minifigure crew members (camera operator, boom mic, director) interacting with film equipment, all rendered with dramatic, high-contrast cinematic lighting.
Prompt Preview
{ "scene_overview": { "setting": "Gothic stone interior resembling a church or cathedral set", "environment_style": "LEGO miniature diorama", "location_type": "Film set / behind-the-scenes movie production", "time_of_day": "Interior, controlled lighting (cinematic night-like mood)", "atmosphere": "Cinematic, eerie, theatrical, controlled", "mood": "Dark humor, spooky, playful, creative", "genre_influence": ["Horror", "Behind-the-scenes filmmaking", "Fantasy"] }, "primary_subject": { "type": "LEGO minifigure character", "role": "Nun / horror character", "appearance": { "outfit": "Black nun habit with white collar", "face": "Pale white face, dark hollow eyes, eerie expression", "accessories": "Cross necklace", "pose": "Standing still, facing forward" }, "positioning": { "location": "Centered within a large stone archway", "background": "Glowing blue-lit alcove or chamber" } }, "supporting_characters": { "crew_members": [ { "role": "Camera operator", "object_interaction": "Operating a large LEGO film camera", "position": "Front-left, seated on a camera dolly" }, { "role": "Boom microphone operator", "object_interaction": "Holding boom mic extended toward subject", "position": "Upper-left foreground" }, { "role": "Lighting technician", "object_interaction": "Holding a rectangular light reflector", "position": "Right side, elevated on a stand" }, { "role": "Clapperboard operator", "object_interaction": "Holding a clapperboard", "position": "Front-center-right" }, { "role": "Director / Script supervisor", "object_interaction": "Holding clipboard and papers", "position": "Bottom-right foreground" } ], "style": "All characters are LEGO minifigures with yellow skin tone and varied hairpieces" }, "props_and_equipment": { "film_equipment": [ "LEGO cinema camera", "Camera dolly on rails", "Boom microphone with fuzzy windscreen", "Lighting reflector panel", "Clapperboard", "Monitor screen displaying the nun character", "Cables coiled on the floor" ], "architectural_elements": [ "Stone brick walls", "Large arched doorway", "Recessed alcove", "Tiled stone floor" ] }, "lighting": { "primary_light": "Strong cyan-blue glow from behind the nun", "secondary_lighting": "Warm neutral studio lighting on crew and foreground", "contrast": "High contrast between cold blue background and warm foreground tones", "style": "Cinematic, dramatic, horror-inspired lighting" }, "color_palette": { "dominant_colors": ["Dark gray", "Black", "Cyan blue"], "accent_colors": ["Warm yellow ("
A Japanese prompt attempting to use Nano Banana Pro to generate a 16-panel grid showing different facial expressions (expression differences) of a heroine, based on a reference image. The prompt specifies that the top-left panel should be the original image, and the remaining 15 panels should show variations while maintaining the art style, similar to pseudo-reference functionality in older models.
Prompt Preview
Create expression differences for the heroine. Display them in a 16-panel screen. The top-left panel should be the original image as is. The remaining 15 panels should display variations of different expressions.
A highly constrained JSON prompt designed to generate a triptych (three vertical panels) of a couple on a swing set, requiring strict identity locking for both subjects and an exact composition lock based on a reference image. The prompt specifies ultra-photorealistic rendering, cinematic lighting (28mm lens, f/2.8), and detailed constraints to ensure the action (upside-down hanging, pushing) and faces are perfectly preserved across all three panels.
Prompt Preview
{ "generation_request": { "meta_data": { "tool": "NanoBanana Pro", "task_type": "photoreal_cinematic_couple_triptych_storyboard", "version": "v1.1_SWING_SET_TRIPTYCH_PLAYFUL_UPSIDE_DOWN_ULTRA_REAL", "priority": "highest" }, "references": { "female_reference_image": { "slot": 1, "purpose": "FEMALE_IDENTITY_LOCK", "strict_identity_lock": true, "face_similarity_priority": "MAX", "no_identity_blending": true, "preserve_age_gender": true, "preserve_skin_texture": true, "preserve_facial_proportions": true, "no_beautify": true, "no_makeup_invention": true }, "male_reference_image": { "slot": 2, "purpose": "MALE_IDENTITY_LOCK", "strict_identity_lock": true, "face_similarity_priority": "MAX", "no_identity_blending": true, "preserve_age_gender": true, "preserve_skin_texture": true, "preserve_facial_proportions": true, "no_beautify": true, "no_makeup_invention": true }, "composition_reference_image": { "source": "UPLOAD_REFERENCE_IMAGE (REQUIRED_FOR_LAYOUT_LOCK)", "purpose": "EXACT_TRIPTYCH_LAYOUT_COMPOSITION_ACTION_LOCK", "strict_lock": true, "no_layout_drift": true, "preserve_framing": true, "preserve_pose_logic": true, "preserve_panel_spacing": true } }, "output_settings": { "aspect_ratio": "4:5", "orientation": "portrait", "resolution": "ultra_high_res", "render_style": "true_photoreal_cinematic_lifestyle", "sharpness": "natural_camera_sharpness_not_ai", "film_grain": "subtle_35mm_analog", "color_grade": "natural_warm_film_neutral_skin", "dynamic_range": "natural_not_hdr", "skin_rendering": "real_microtexture_pores_peachfuzz_no_plastic", "compression_artifacts": "none" }, "hard_constraints": [ "Exactly 2 people only (one female, one male).", "Faces must match the uploaded references with maximum similarity (no identity drift).", "Triptych layout: THREE horizontal panels stacked vertically in one 4:5 image, exactly like the composition reference.", "Outdoor playground swing set setting, same action beats as reference: playful upside-down hanging + pushing + reaching toward camera.", "Faces must be clearly visible and recognizable in ALL THREE panels (no hair fully covering face, no extreme blur).", "Photorealistic anatomy; correct hands/fingers; realistic clothing folds and gravity.", "Swing set geometry must be physically correct (straight metal frame, correct chain links, no warping).", "No text, no logos, no watermark." ], "camera": { "style": "real_cinematic_lifestyle_photo", "lens": "28mm_full_frame", "distortion": "realistic_28mm_mild_edge_distortion_controlled", "aperture": "f/2.8" } } }
A simple, creative prompt structure for generating an image of a massive kaiju kitten destroying a specific urban landmark, viewed from a ground-level perspective under bright daylight.
Prompt Preview
Massive kaiju kitten destroying Sydney's urban waterfront, destroying the {argument name="landmark name" default="[place landmark name here]"}. Ground-level view. Bright, daytime, blue skies. No glowing eyes.
A complex, surreal prompt to generate a hyper-realistic, cinematic image of a colossal Taylor Swift standing in a city street, surrounded by construction scaffolding and miniature construction workers, emphasizing scale and urban fantasy aesthetics.
Prompt Preview
"A colossal woman standing in a city street, appearing to be 'under construction' with scaffolding surrounding her body.", "aesthetic": "Hyper-realistic composite, cinematic, urban fantasy." }, "subject_details": { "celebrity_likeness": "{argument name="celebrity name" default="Taylor Swift"}.", "features": { "face": "Blue eyes, signature red lipstick, distinct facial structure.", "hair": "Golden blonde, long, wavy texture cascading over shoulders." }, "apparel": { "dress": "Black satin slip dress with a deep V-neck and a high slit on the right thigh.", "accessories": "Holding a standard white disposable coffee cup in the right hand. Carrying a brown leather luxury handbag in the left hand.", "footwear": "Bare feet or heels (obscured by street debris/cracks)." } }, "surreal_elements": { "scaffolding": "Intricate metal construction scaffolding structures built around her arms, torso, and legs, following the contour of her body.", "miniatures": "Tiny construction workers in yellow hard hats and high-vis vests climbing the scaffolding on her body.", "destruction": "The asphalt street beneath her feet is cracking and shattering under the immense weight, creating a crater pattern." }, "exact_pose_anatomical": { "posture": "Standing tall with a slight contrapposto (weight shifted to one hip).", "orientation": "Front-facing, towering over the viewer/city.", "right_arm": "Bent at the elbow, raised to hold the coffee cup at chest level.", "left_arm": "Extended downwards holding the handbag.", "gaze": "Direct, calm eye contact with the camera.", "scale": "Subject appears to be approximately 50 stories tall, matching the height of surrounding skyscrapers." }, "environment": { "location": "Avenue in a major metropolis (resembling Chicago or NYC).", "elements": "Skyscrapers flanking both sides, busy traffic with miniature yellow taxis and buses way below.", "perspective": "Aerial view relative to the city, but eye-level relative to the subject." }, "camera_technical_values": { "lens": "35mm (Wide angle to capture the full scale of the subject against the skyline).", "aperture": "f/11 (Deep depth of field to keep both the giant subject and the city background relatively sharp).", "shutter_speed": "1/500s.", "iso": "ISO 100.", "resolution": "8k, high fidelity textures." }, "lighting_setup": { "source": "Natural Daylight.", "quality": "Soft, diffused sunlight (slightly overcast).", "direction": "Overhead/Frontal.", "shadows": "Subject casts a massive, long shadow onto the city streets below."
A prompt designed to generate non-photorealistic, high-energy comic book art. It specifies bold outlines, vibrant colors, dramatic action, and explicitly excludes photos or realism, aiming for a pure, print-ready comic style.
Prompt Preview
High-energy panels, bold outlines, vibrant colors, dramatic action. No photos. No realism. Just pure, print-ready comic art from one Nano Banana Pro prompt.
A prompt for Nano Banana Pro to create a hyper-realistic, surreal composite image using forced perspective. It features an extreme close-up of a man's astonished face and hand, cradling a tiny, miniature version of a man, emphasizing 8k Octane Render fidelity and skin texture.
Prompt Preview
A hyper-realistic, surreal composite shot utilizing forced perspective to create a dramatic scale difference, featuring an extreme close-up profile view on the right side of a man with uploaded face as reference, his expression one of slack-jawed astonishment with mouth slightly agape as he gazes downward. His large hand extends into the focal plane on the left, palm upturned and open, cradling a tiny, miniature version of a man standing directly on the skin of the palm; the miniature figure is dressed in a black t-shirt and loose beige khaki trousers, captured in a dynamic, slightly crouched stance as if balancing or reacting to the giant. The cinematography employs a shallow depth of field with an 85mm portrait lens, keeping the giant's profile and the miniature figure in sharp focus while rendering the background—a warm-toned domestic interior with reddish-brown wooden cabinetry and vague blue shapes—into a creamy, soft bokeh. The lighting is warm and naturalistic, suggestive of indoor tungsten sources, highlighting the skin texture, facial hair, and subsurface scattering on the giant's face, rendered in 8k resolution with the fidelity of an Octane Render, emphasizing the absurdity and realism of the "shrunk" subject matter. Man should be fully visible
No prompts found for the selected type.
Try adjusting your search or filters