50 prompts picked from the 150 in YouArt's awesome-gpt-image-2-5-prompts, grouped by what you would make with them: portraits, product campaigns, typography posters, cinematic scenes, storyboards, focused edits and character sheets. 12 are JSON, 38 are prose. Each prompt is shown verbatim with its author, followed by my notes on why it works, which parts carry the effect, and what to change to reuse it. At the end, a style library lists 103 restyle prompts from Visual AI Club (LEGO, Pixar-style 3D, game art, toys, caricatures, outfits) with their example images.
Ronald Luc · 2026-09-26 · prompts from the repo snapshot of 2026-09-10 · 60 images generated for this page with GPT Image 2.5
Scroll the covers in the style library and open the carousels you like.
Read properly (about 75 minutes, plus 30 for the style library)
Open the prompts and find the parts the notes point at.
Compare the pairs: each JSON example links to a prose one that does the same kind of job.
Copy a prompt, change the parts listed under “To reuse”, and run it.
Card anatomy
JSON or Prose: the prompt format. Labeled sections without braces count as prose.
Edit: needs an input image: the prompt refers to an uploaded image and makes little sense without one.
Template: the prompt has slots. {argument name="x" default="y"} is YouMind's slot syntax; replace the whole token with your value. Slots are highlighted in yellow in the prompt text.
Suggested model: Sunburst where text, faces, edits or exact products matter; “Either” where Flare is likely enough.
Images: generated for this page through the OpenAI Images API, which sends the prompt as written. Each caption gives the model, quality, size and job number, and says what the model got right and where it did not follow the prompt. Any change from the verbatim prompt is listed on the card. Every card has at least one image; edits show their input image as a small thumbnail.
Attempts: one sample per prompt, no cherry-picking, with these exceptions. J13 and J43 are second attempts: the first ones had a real space-agency logo on a suit and a real beer brand's sign on a building. J53 was retried once because a real magazine title appeared on a book spine; the retry had it too, so the first attempt is shown. The verbatim prompts for J03, J04, J05 and J50 were refused twice by 2.5 output moderation. J03, J05 and J50 come from minimally edited prompts, and their cards show the edit. J04, the Flare copy of J03, was refused on every version and has no image. J25-J28 are input images I made from my own prompts.
The Copy button always copies the original prompt text, even where the display adds line breaks.
What carries over to 2.5
Most of these prompts were written for GPT Image 2 before the 2.5 models shipped on 2026-09-08. They carry over. The short version of my notes on 2.5:
Two models.gpt-image-2.5-flare is the small, fast one and the default in most apps. gpt-image-2.5-sunburst is the quality one. Single shots are close; Sunburst leads Flare by 44 Arena points on edits. Sunburst keeps pores and flyaway hairs; Flare smooths skin and warms the grade. Draft on Flare, finish on Sunburst.
Where you send the prompt matters. ChatGPT and the Responses API image tool rewrite your prompt, JSON included, before the image model sees it. The Images API sends it as written. Test prompts in the channel you will use.
No seed. Output varies a lot even with a detailed prompt. Generate 4-8 and pick, rather than rewriting after one bad sample.
Size is a parameter, not a word. Edges are multiples of 16, the long edge is at most 3840 px, the ratio at most 3:1, and anything above 2560x1440 is experimental. “8K”, “--ar 4:5” or "aspect_ratio" inside the prompt are read as text.
Describe physical facts, not praise. A capture device, named light sources, 2-3 concrete imperfections. Drop “8K, masterpiece, hyper-detailed”; they push towards the over-sharpened look.
Text: quote the exact words, say where and in what typography, count the elements, add “no extra text”.
Counts: “exactly 5”, “1 citrus wedge”. Anything that can multiply should have a number.
Edits: change one thing, repeat the full preserve list every turn, and stop chaining after 2-4 rounds. 2.5 drifts much less than 2.0 (18% vs 60% of pixels moved after five edit rounds in one test), but it is still not pixel-exact.
No memory between calls. A prompt cannot create a stored identity or remember earlier outputs. Consistency comes from reference images and fixed wording.
Beautification bias. All current models make faces smoother, younger and more symmetrical (arXiv 2608.02841). Say what to keep: freckles, pores, apparent age.
The six common mistakes, from @1littlecoder's guide: adjective soup, naming a style instead of describing it, vague euphemism, text without typography, shrinking the preserve list during edits, and batching several changes into one turn. Several prompts below show one of them; the notes say where.
Credits and licences
Prompts are by their authors. Every card names the author and links the original post. The prompt text is quoted verbatim; changes made for the generated images are listed on the card.
Collection:youart-open-source/awesome-gpt-image-2-5-prompts by YouArt, commit f7f6276, corpus of 2026-09-10 (150 prompts, 110 authors). The repo's selection, titles and descriptions have no licence grant, so the titles, grouping and notes here are my own.
Upstream terms. YouArt took the prompts from two collections and records the terms each collection stated for its copy. That is a record of what was received, not a licence from the authors.
10 prompts came via YouMind-OpenLab/awesome-gpt-image-2 (commit 4157207, 2026-09-09), which stated CC BY 4.0 with this notice: “Copyright (c) 2026 YouMind OpenLab. Licensed under Creative Commons Attribution 4.0 International (CC BY 4.0), https://creativecommons.org/licenses/by/4.0/. No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.”
Example images in the repo are not reproduced here: the repo grants no licence for them. The section headers link to them on GitHub instead.
Images on this page, except in the style library, are generated by me with GPT Image 2.5 from the prompts shown. All people in them are fictional. The edits of a person start from a fictional person generated from my own prompt (J01); the product and crowd edits start from a packshot or photo I generated (J26, J27, J28).
Style library. The prompts and example images in the style library are from visualaiclub.com and belong to their authors. Each card names the author and the licence chosen on the site, and links to the source page.
If you wrote one of these prompts and want it removed or credited differently, open an issue on the blog repo.
Your list vs the repo
The repo sorts its 150 prompts into 9 categories (about 16 each): portrait, ecommerce-product, ad-creative, poster, typography-text, illustration, character-design, infographic and ui-mockup. The groups on this page follow what the prompt makes, not the repo category, so most groups draw from several categories. “Board focus edit” I read as “focused edit”, the README's label for image-to-image changes; storyboards got their own group.
gpt-image-2.5-sunburst · high · 1024x1280 · job J02Result: Matches the prompt: tilted frame, flare, blur, grain and a degraded face, with the teacup and book clipped in the foreground.
Works because. It names one capture device, a 2000s 0.3-megapixel toy digital camera, and then lists what that device does: motion smear, pop-up flash in daylight, JPEG damage, chromatic aberration. The model gets physical facts, not a mood.
Carries the effect. “Do not protect facial sharpness” and “compression damage visible across the face”. Without them the model keeps the face clean while everything else degrades, which is exactly what looks fake.
To reuse. Swap the first paragraph (scene) and keep the rest as a block. --ar 4:5 is Midjourney syntax; on the API it is just text, and the size parameter sets the frame.
Watch out. The 17-term “Negative prompt” line is read as plain text; GPT Image has no negative-prompt parameter. My notes recommend 2-3 exclusions aimed at the tells you actually see.
Sent for the image: Wardrobe line changed: “a simple orange bralette” became “a simple white cotton tank top”. Everything else verbatim.
Prompt, verbatim · 2,713 characters · 388 words
Show the full prose prompt
A quiet summer morning in a European apartment. A woman sits on the floor next to open balcony doors, absorbed in a book, one leg stretched toward the railing, a cream knitted blanket draped loosely over her lap. She wears a simple orange bralette and delicate jewelry, completely unaware of the camera. A cup of tea and an open book rest on the wooden floor beside her. The balcony overlooks classic European architecture: cream stone facades, wrought-iron railings, flowering balconies, leafy trees.
Photographed from an awkward handheld angle, as if someone walked past the room and snapped a photo without slowing down. The framing is tilted and accidental. The subject is not centered. Parts of the balcony doors, blanket, floor, and foreground objects are cut off by the frame edges.
Captured on a cheap 2000s low-resolution toy digital camera. 0.3-megapixel quality. Heavy motion blur. Intentionally crooked and tilted framing. Harsh pop-up flash despite bright daylight. Overexposed white highlights. Intense digital noise and heavy sensor grain. Low-fidelity faded colors. Chromatic aberration at the edges. Lens flare. Looks like an extremely blurry candid photo uploaded from an old toy camera. Raw and amateurish. Authentic nostalgic texture.
Slight focus miss, motion smear, low-resolution facial details, sensor noise, JPEG artifacts, and compression damage visible across the face — exactly as a cheap toy camera would produce.
Do not preserve modern clarity. Do not protect facial sharpness. The subject remains recognizable despite image degradation. Subject occupies 65% of the frame.
Large foreground obstructions from a partially clipped teacup, book, blanket folds, and balcony door frame. Extreme off-balance composition. Camera not leveled. Subject positioned awkwardly near one edge. Parts of the room, railing, and foreground elements are clipped.
The image should feel like a forgotten candid uploaded to an old social media site in the mid-2000s. Raw, amateurish, imperfect, accidental, nostalgic, unpolished. Captured-memory realism, not aesthetic photography.
Super thick plastic toy-lens softness, motion smear, focus miss, smeared highlights, low-resolution detail loss, heavy JPEG compression, sensor noise, and degraded detail across the face, hair, hands, blanket, books, teacup, balcony, buildings, and sky.
--ar 4:5
Negative prompt: beauty filter, modern smartphone quality, DSLR quality, cinematic grading, HDR photography, professional composition, studio lighting, ultra sharp face, perfect skin, flawless focus, fashion editorial, influencer photo, polished digital rendering, AI perfection, hyper-detailed eyes, premium camera quality, luxury portrait photography
gpt-image-2.5-sunburst · high · 1024x1536 · job J03Result: Close to the rewritten prompt. The neckline came out lower than “modest”, and the frame runs down to the thighs instead of cropping near the denim waist.
Works because. Each section answers one question: who, wearing what, lit how, framed where. The composition block puts the face, the hand and the window at named places in the frame.
Carries the effect. The light has a source, a side and a hardness: “somewhat harsh warm primary light shines from the upper left window, creating thin window frame shadows”. The negative line guards the thing models flip most often, the light direction.
To reuse. Keep Background/Lighting and Composition/Camera, change Main Subject and Clothing. The {argument ...} tokens are YouMind's template slots; replace the whole token with plain text.
Sent for the image: Template slots filled with their defaults.
Edited to pass 2.5 moderation. As written, this prompt was refused twice by GPT Image 2.5 output moderation in my run on 2026-09-26, on Sunburst (J03) and on Flare (J04). The image is Sunburst's result for the second of three minimal rewrites. Changes (removed › added):
glossy lips slightly open › lips closed
a deep V-neck › a modest scoop neck
draped off both shoulders › worn on the shoulders
light blue denim shorts › light blue denim jeans
do not have the cardigan fully worn up to the shoulders. › do not have the cardigan slipping off the shoulders.
The first rewrite, which had every change except the lips, was refused. So was a third that also deleted the chest tie, so the filter is not consistent. Flare refused the exact version Sunburst accepted and 3 further rewrites (dropping “sheer”, the skin “gloss” and the camisole straps), so there is no Flare image. See Flare vs Sunburst.
Subject:
Window light and cream-colored lace
Main Subject:
A vertical waist-up photograph of an adult woman wearing a {argument name="top" default="cream-colored lace top"} and {argument name="bottom" default="pale denim"} sitting on a wooden chair by a bright window. The subject occupies a large portion of the center of the frame, with sunlight streaming from the left window and a wooden table in the back right background.
Character/Expression:
Face tilted slightly to the left, looking directly at the camera with a gentle expression and glossy lips slightly open. Slender oval face, small chin, large grayish-brown almond eyes, thin eyebrows, smooth nose bridge, and pale rose-colored lips. Glossy ash-brown hair down to the chest with thin bangs and thin strands along the cheeks, flowing straight from the right shoulder to the chest. A small round pendant.
Clothing/Pose:
A cream-colored fine-ribbed camisole top with wide lace shoulder straps, a deep V-neck, a thin tie in the center of the chest, and lace at the hem. A sheer long-sleeved cardigan of the same color draped off both shoulders, with light blue denim shorts underneath. Sitting in a chair, right elbow raised with the back of the right hand placed under the cheek, and left arm lowered towards the knees.
Background/Lighting:
A white-framed window and white wall on the left, with a simple wooden table and chair in the right background. A somewhat harsh warm primary light shines from the upper left window, creating thin window frame shadows and bright highlights on the face, chest, and sleeves, while leaving soft shadows on the right side.
Composition/Camera:
2:3 vertical composition. The camera captures the upper body from a short distance at approximately the same height as the seated person's eye level, cropped at the top of the head and near the denim waist. Place the face in the top center, the right hand on the cheek in the top right, and the window on the left edge, making the subject large in the center. Focus on both eyes, with the indoor background weakly and naturally blurred.
Texture/Style:
Photorealistic high-definition live-action photograph. Precisely depicts natural skin texture and gloss, individual hair strands, fine ribs, lace, sheer sleeves, and denim weave, adjusted to a warm, low-saturation palette of cream and pale wood colors.
Negative:
Do not reverse the direction of the window light and shadows; do not have the cardigan fully worn up to the shoulders.
gpt-image-2.5-sunburst · high · 1024x1536 · job J29Result: Matches: cross-legged on the balcony, hand on cheek, railing and pink flowers. The feet are cut off at the bottom edge.
Works because. “candid smartphone photo”, “slightly imperfect smartphone-camera details” and “no beauty-filter effect” push away from the glossy default look.
Watch out. The \n sequences are literal backslash-n characters in the source, not line breaks. Replace them before sending. “a Balcony slightly pouty look” is a leftover word from editing.
Contradiction. “Ultra-realistic photography ... 4K realism” sits next to “slightly imperfect smartphone-camera details”. Drop the resolution words; they pull towards the over-sharpened look.
To reuse. The three slots (subject, hair style, clothing) are the variable part. The light and capture sentences are the fixed part.
Sent for the image: Template slots filled with their defaults, and the literal \n sequences turned into line breaks.
Prompt, verbatim · 2,438 characters · 320 words
Show the full prose prompt
A completely natural, photorealistic candid smartphone photo of a {argument name="subject" default="young East Asian woman"} sitting comfortably cross-legged on a balcony garden terrace. She has {argument name="hair style" default="long, naturally wavy dark brown hair"} falling loosely over her shoulders and down the front of her body, with soft individual strands moving naturally. Her features are delicate and realistic, with subtle natural makeup and authentic skin texture.\nShe is wearing an {argument name="clothing" default="oversized cozy light-gray sweatshirt"} with relaxed-fit light-blue jeans. Sitting casually and naturally, she gently rests one hand against her cheek while looking upward and slightly off to the side. Her expression feels genuinely cute and thoughtful, with a Balcony slightly pouty look, as if the photo was captured during a quiet unposed moment.\nShe sits on a light-colored outdoor terrace surface surrounded by lush green plants, leafy potted foliage, and small blooming pink flowers. A simple black metal balcony railing runs behind her, with dense trees and natural greenery filling the background. The setting feels peaceful, private, and like a real everyday garden balcony.\nSoft warm natural daylight falls gently across her face and clothing, creating realistic subtle shadows. Natural skin pores, authentic hair texture, realistic sweatshirt fabric, and believable denim details are clearly visible. Nothing looks overly polished or artificially posed.\nShot as an authentic casual smartphone photograph with a natural eye-level perspective. Full-body seated composition, vertical framing, with the subject positioned slightly toward the lower-middle of the frame. The background has gentle natural depth of field—softly blurred while remaining clearly recognizable.\nUltra-realistic photography, natural proportions, candid everyday moment, Korean-inspired casual fashion aesthetic, muted soft colors, subtle film grain, realistic lighting, slightly imperfect smartphone-camera details, 4K realism, no beauty-filter effect, no excessive retouching.\nNegative prompt: cartoon, anime, illustration, CGI, artificial-looking skin, plastic skin, excessive makeup, distorted facial features, extra fingers, malformed hands, extra limbs, incorrect anatomy, unnatural body proportions, stiff pose, oversaturated colors, dramatic studio lighting, blurry face, low resolution, watermark, text, logo.
gpt-image-2.5-sunburst · high · 1536x864 · job J05Result: All 8 labels as written, 3 focus circles and one cyan fill-light arrow. Focus Point 1 is on the knees, as the rewrite asks.
Works because. Two layers are described separately: the photo (subject on the left third, a bright window mass on the right) and the overlay (grid, markers, arrows, labels), each with colour and position.
Carries the effect. The eight overlay labels are quoted exactly, one per line. Quoted strings are what makes text come out as written.
To reuse. Works for any composition lesson. Keep the overlay block, change the photo block, and check that every label points at something the photo actually contains.
Check. Count the result: 3 focus circles, 8 labels, one cyan fill-light arrow.
Edited to pass 2.5 moderation. As written, this prompt was refused twice by GPT Image 2.5 Sunburst output moderation in my run on 2026-09-26. The image comes from the fourth of four minimal rewrites, the first one that passed. Changes (removed › added):
kneeling on a soft white bed › wearing a cream knit sweater and linen trousers, sitting on the edge of a soft white bed
Her body leans slightly backward, facing › Her body turns slightly, facing
Waist and hip area as Focus Point 1. › Knees and bedding as Focus Point 1.
from the face to the waist/hip, › from the face to the knees,
face, upper body, waist, and silhouette › face, upper body, and silhouette
Rewrites 1-3 built up the same list one step at a time and were all refused. Only the fourth, which also adds clothing (the original prompt describes none), passed.
The prompt below is the verbatim original.
Prompt, verbatim · 2,428 characters · 385 words
Show the full prose prompt
Create a cinematic fashion editorial photograph in a bright minimal bedroom, presented as a cinematography composition analysis frame.
Main Image: A young adult female model kneeling on a soft white bed beside a large window with sheer white curtains. She is positioned on the left third of the frame, body aligned with the vertical rule-of-thirds grid line. Her body leans slightly backward, facing toward the glowing window light with a calm, distant, introspective expression. The right side of the image is filled with a large bright white luminous mass from overexposed daylight through sheer curtains, balancing the darker subject on the left.
Lighting: Strong natural window light from the right side creates high contrast on the model's face, upper body, waist, and silhouette. Add subtle secondary fill light from the left side to reveal shadow areas without flattening the contrast. Warm hazy morning atmosphere, soft highlights, realistic skin texture, gentle shadows, soft bloom near the curtains.
Composition Overlay: Add visible cinematography composition guide graphics over the image, like a film school analysis frame.
Include:
- Rule of thirds grid lines in thin green.
- A red vertical line showing the model's body aligned on the left rule of thirds grid.
- Green circular focus markers around:
1. Waist and hip area as Focus Point 1.
2. Face area as Focus Point 2.
3. Bright curtain/window mass as Focus Point 3.
- Yellow eye-trace arrows moving from the face to the waist/hip, then toward the glowing window.
- Red leading-line arrows following the bed edge and wall panel lines toward the main focus point.
- A cyan arrow showing secondary fill light coming from the left side.
- Text labels placed cleanly around the frame:
“Body aligned on the grid of thirds”
“High contrast Focus Point 2”
“Eye Trace”
“Focus Point 3”
“High contrast Focus Point 1”
“Secondary light source for dark parts”
“Lines lead to the main focus point”
“The white mass is aligned on the grid of thirds to balance the left side of the image”
Style: Luxury fashion editorial photography, soft cinematic realism, warm beige and white color palette, minimal bedroom, cream wall panels, white bedding, sheer curtains, subtle film grain, shallow depth of field, elegant composition, educational cinematography breakdown overlay.
Camera: 35mm lens, medium wide framing, eye level camera angle, 16:9 widescreen.
gpt-image-2.5-sunburst · high · 1536x1536 · job J30Result: 16 cells, faces blurred as asked. The waving cell (row 3, left) shows an open palm toward the lens.
Works because. The images array gives each of the 16 cells a position and a one-line shot description. That is the part prose handles badly: 16 distinct frames you can address one by one.
Carries the effect."faces": "intentionally obscured by soft rectangular blur blocks". A privacy device that also avoids the problem of keeping two faces consistent across 16 frames.
To reuse. Keep style, camera_look and color_palette, rewrite the 16 descriptions. For the same people in every cell you still need a reference image; JSON does not store an identity.
Sent for the image: The two peace signs changed (I keep hand gestures out of the images here): “giving a peace sign” became “waving at the camera”, “one subject making a peace sign” became “one subject laughing”. Everything else verbatim.
Prompt, verbatim · 3,812 characters · 456 words
Show the full JSON prompt
Line breaks added for reading. Copy gives the original single line.
{
"type": "16-photo nostalgic contact sheet collage",
"style": "dreamy film photography, soft blur, slightly underexposed, candid youthful romance, flash snapshots mixed with ambient dusk light, subtle grain, sentimental and bittersweet mood",
"subject": {
"people_count": 2,
"relationship": "young couple or former lovers spending time together",
"ages": "early 20s",
"appearance": {
"male": {
"build": "slim",
"hair": "short dark hair",
"clothing": "loose white short-sleeve shirt, camera strap around neck in several shots"
},
"female": {
"build": "slim",
"hair": "shoulder-length dark hair",
"clothing": "light sleeveless tops or soft casual summer clothes"
}
},
"faces": "intentionally obscured by soft rectangular blur blocks over every visible face"
},
"layout": {
"grid": {
"rows": 4,
"columns": 4,
"count": 16,
"border": "thin white dividers, equal square cells"
},
"images": [
{
"position": "row 1 col 1",
"description": "close cropped portrait of the woman in a white top at night, soft flash, dark background"
},
{
"position": "row 1 col 2",
"description": "close cropped blurred two-person selfie framing, both subjects partially visible, dark nighttime setting"
},
{
"position": "row 1 col 3",
"description": "young man standing at night and holding a compact silver camera up to his face, white shirt, distant lights behind him"
},
{
"position": "row 1 col 4",
"description": "woman on a beach or shoreline in low light, softly blurred, ocean horizon behind her"
},
{
"position": "row 2 col 1",
"description": "street candid of the man holding a camera near his face while walking outdoors in the evening, urban background with motion blur"
},
{
"position": "row 2 col 2",
"description": "close-up of the woman indoors or in a dim warm setting, hand raised near her face, flash-lit snapshot"
},
{
"position": "row 2 col 3",
"description": "blurred two-shot of the couple sitting close together by water at dusk, intimate candid composition"
},
{
"position": "row 2 col 4",
"description": "young man outdoors in greenery during daytime or early evening, looking down at a camera in his hands, white shirt and camera strap visible"
},
{
"position": "row 3 col 1",
"description": "woman close to the camera giving a peace sign, casual sleeveless top, sandy or beachlike background"
},
{
"position": "row 3 col 2",
"description": "back view of the man in a white shirt looking out over a cityscape at night from a high vantage point"
},
{
"position": "row 3 col 3",
"description": "woman indoors at night holding a compact camera directly toward the viewer, city lights beyond a window, flash aesthetic"
},
{
"position": "row 3 col 4",
"description": "tight cropped two-person selfie-like frame with both subjects partially visible, dark background"
},
{
"position": "row 4 col 1",
"description": "young man at the waterfront at dusk holding a camera to his eye, cloudy blue sky and distant shoreline behind him"
},
{
"position": "row 4 col 2",
"description": "soft night portrait of the woman on a city street with warm bokeh lights in the background"
},
{
"position": "row 4 col 3",
"description": "close intimate couple snapshot with both faces near each other, one subject making a peace sign, heavy blur and flash look"
},
{
"position": "row 4 col 4",
"description": "rear view of the woman walking alone down a warmly lit narrow street at night, shoulder-length hair and light top visible"
}
]
},
"composition": "each square feels like a memory fragment from one summer evening and a few nearby outings, varied framing, natural imperfection, casual amateur photography",
"color_palette": "muted blues, warm tungsten yellows, soft skin tones, dark greens, charcoal night shadows, faded white clothing",
"camera_look": "35mm point-and-shoot or disposable camera feel, shallow focus, motion blur, bloom around lights, occasional flash overexposure",
"quality": "high-resolution collage with authentic analog softness, emotionally evocative and realistic"
}
gpt-image-2.5-flare · medium · 1024x1536 · job J31Result: The adjective soup shows: glossy, over-sharpened, beautified face. Not asked for: “THE NEXT CHAPTER”, “STRENGTH BEAUTY LEGACY”, “AN ETERNAL BALANCE” and a red seal.
What goes wrong. The last paragraph is about 20 quality phrases: “32k detail”, “editorial masterpiece”, “advanced ray tracing”, “photorealistic perfection”. None is a visual fact. This is the “adjective soup” mistake, and these words push towards the over-sharpened AI look.
Mixed signals. “visible skin pores, subtle freckles” (good) sits next to “perfectly balanced proportions” and “premium salon-quality shine” (beautification). The model gets both instructions and averages them.
Keep. The first two paragraphs and the camera paragraph (110mm, large diffused key light, subtle rim light) are concrete.
Fix. Delete the last paragraph, cut “natural beauty, perfectly balanced proportions”, add “neutral white balance”.
Sent for the image: Verbatim, adjective soup and all. The image is the lesson.
Prompt, verbatim · 2,854 characters · 342 words
Show the full prose prompt
Ultra-photorealistic cinematic portrait of a legendary futuristic martial arts heroine standing in the center of a luxury fashion-editorial composition. She wears an elegant crimson and black combat outfit inspired by contemporary haute couture and advanced athletic armor, featuring intricate embroidered patterns, premium silk textures, carbon-fiber accents, metallic details, and subtle illuminated elements integrated into the design.
The character is captured in a dynamic low stance, one hand extended forward and the other resting near her waist, projecting calm confidence and immense strength. Her facial features are exceptionally realistic with natural beauty, perfectly balanced proportions, visible skin pores, subtle freckles, fine facial hairs, realistic skin translucency, delicate imperfections, soft blush tones, detailed lips with natural moisture, and ultra-sharp eye detail. Deep hazel eyes display complex iris structures, realistic reflections, and cinematic catchlights.
Her long dark hair flows naturally through the frame, with thousands of individually rendered strands, realistic flyaway hairs, subtle movement, and premium salon-quality shine. Decorative fabric ornaments and metallic accessories add visual interest while maintaining realism.
The composition incorporates an artistic mixed-media gallery backdrop featuring fragmented photographic panels, oversized close-up facial details, layered fashion-magazine elements, torn-edge textures, floating geometric shapes, translucent acrylic sheets, and contemporary luxury-brand advertising aesthetics. The background remains predominantly bright and minimalist with sophisticated negative space.
Wardrobe materials showcase extraordinary realism including premium satin, brushed metal, woven fabric, polished leather, textured embroidery, reflective surfaces, and physically accurate material responses. Every fold, stitch, seam, wrinkle, and fabric tension point is visible with microscopic precision.
Captured using a medium-format professional camera system, 110mm portrait lens, ultra-high dynamic range imaging, shallow depth of field, razor-sharp focus on the eyes, cinematic studio lighting with a large diffused key light, subtle rim illumination, controlled fill lighting, realistic shadow transitions, luxury fashion photography quality, and premium commercial advertising aesthetics.
Hyper-detailed skin rendering, subsurface scattering, realistic global illumination, advanced ray tracing, physically based rendering, volumetric atmosphere, ultra-clean composition, museum-quality portrait photography, luxury campaign aesthetics, editorial masterpiece, award-winning fashion photography, extraordinary realism, 32k detail, ultra-high fidelity textures, cinematic depth, breathtaking realism, next-generation rendering quality, photorealistic perfection.
Product campaigns
Seven product prompts, from a JSON ingredient ad to a jewelry prompt that needs your own product photo. The recurring tools are exact label text, counted props and a product lock.
YouArt's own example image for this category, on GitHub (not copied here): product.webp
gpt-image-2.5-sunburst · high · 1152x1536 · job J06Result: All four labels as written, including the duplicated brownie label on the vanilla side.
Works because. The nesting follows what a food ad needs: product, layers, stick, then callouts, each with exact text, a position and a visual aid. A layout with four labels is where structure helps.
Bug in the source. “Chocolate ice cream with brownies” is used twice, left-middle and right-middle, while the right side of the bar is vanilla with caramel. The model draws what it is told. The same bug in prose gives the same result; see JSON vs prose.
Does nothing."resolution": "8K" and "aspect_ratio": "3:4" are text, not settings. The size parameter decides the frame.
To reuse. Replace ice_cream_bar and labels; keep lighting, background and the callout style.
gpt-image-2.5-sunburst · high · 1024x1280 · job J32Result: Headline, supporting line and $5.80 badge as written and horizontal. The full stop in “Bliss.” was rendered, as the notes predicted. The metal cone holder looks odd.
Works because. Short (1,233 characters) and ordered: product, background, typography, light, mood. Each text element has its own line and a role: headline, supporting line, badge.
Carries the effect. “(not vertical)”: a small fix for a common failure where the model stacks the text vertically.
Watch out. The headline is written as “Sweet Strawberry Bliss.” with a full stop, so you may get the full stop. Quote strings without punctuation you don't want, and add “no extra text”.
Ultra-realistic product photography of a rich strawberry soft-serve ice cream in a crispy waffle cone, styled with a clean, modern premium aesthetic. The soft serve is a vibrant natural pink, thick and creamy, sculpted into a smooth swirl with a softly curled peak, lightly topped with delicate strawberry dust or tiny fruit specks for a fresh, appetizing look. The cone has a rustic, crunchy texture with slightly uneven edges for an artisanal feel.
The background is soft beige with natural sunlight casting subtle leaf shadows, creating a calm, organic atmosphere. Include softly blurred greenery in the foreground for depth. The composition is minimal, balanced, and uses negative space effectively, similar to high-end American food brand ads.
On the left side, include modern English typography in a clean, elegant layout (not vertical).
Main headline:
Sweet Strawberry Bliss.
Supporting line (smaller text):
Made with real strawberries. Smooth. Creamy. Irresistible.
Add a small circular badge showing the price:
$5.80.
Lighting: soft natural daylight, warm highlights, shallow depth of field, high-end commercial food photography style.
Mood: fresh, premium, modern, and inviting — aligned with upscale U.S. dessert branding.
gpt-image-2.5-sunburst · high · 1024x1280 · job J08Result: Label text correct. Not asked for: a line under the logo (“NATURAL BEAUTY FOR A BRIGHTER YOU”) and two slogans on the set (“NATURE NOURISHES A BRIGHTER YOU”, “PURE PLANTS BRIGHTER SKIN”).
Works because. The label content is given as an ordered list (“AUREVIA, VELVET HYDRATE, Nourishing Body Lotion, Aloe Vera + Shea Butter, Soften • Nourish • Hydrate, 250 ml”), and the bottle is described in construction terms: 250 ml tall oval pump bottle, soft-touch finish, brushed champagne-gold pump.
Leftovers. “Avoid copying the rectangular label ... of the reference” and “Instead of an orange-toned surface” refer to a reference image that is not sent here. The prompt was written as an edit. Without the reference it still works; the leftovers just say what not to do.
To reuse with a real product. Send the product photo as image 1 and add a protection list: silhouette, label layout, logo, cap and pump shape, colours. The model does not guarantee product identity. For an exact label, paste the real one back in afterwards.
Sent for the image: Template slots filled with their defaults (AUREVIA, VELVET HYDRATE Body Lotion, matte sage-green).
Create an ultra-realistic luxury commercial product advertisement for a premium skincare brand called {argument name="brand name" default="AUREVIA"}, featuring its “{argument name="product name" default="VELVET HYDRATE Body Lotion"}.” Place a sophisticated 250 ml tall oval pump bottle prominently in the center of the composition, positioned at a slight diagonal angle for a dynamic editorial look. Use a {argument name="bottle color" default="matte sage-green"} bottle with a smooth soft-touch finish and a brushed champagne-gold pump. Design the label with an elegant arched botanical pattern inspired by flowing leaves and subtle topographic lines. Use a sophisticated palette of sage green, champagne gold, warm ivory, and muted terracotta. Avoid copying the rectangular label or orange-and-purple styling of the reference. The front label should clearly and correctly display: AUREVIA, VELVET HYDRATE, Nourishing Body Lotion, Aloe Vera + Shea Butter, Soften • Nourish • Hydrate, 250 ml. Surround the bottle with beautifully arranged fresh aloe vera slices, small ivory flowers, shea nuts, eucalyptus leaves, and delicate botanical elements. Add transparent water droplets and a graceful splash of water sweeping behind the bottle to communicate freshness and hydration. Instead of an orange-toned surface, place the product on a light cream travertine stone platform with subtle natural texture. Introduce an artistic curved sage-green backdrop behind the product, creating layered depth and a distinctive premium set design. Use soft golden daylight entering diagonally from the upper left, creating elegant botanical shadows across the background. Add realistic reflections, tiny condensation droplets on the bottle, subtle highlights on the gold pump, and soft shadows beneath the product. Composition should feel sophisticated, fresh, botanical, modern, and expensive—similar to a high-end international skincare campaign. Photography: luxury beauty advertising, professional product photography, macro-level material detail, realistic liquid and water physics, crisp typography, photorealistic botanical textures, shallow depth of field, premium studio lighting, 8K detail. Aspect ratio: 4:5 portrait.
1. Editorial · gpt-image-2.5-sunburst · high · 1024x1280 · job J33Result: Every requested string is correct. Not asked for: “SPARKLING BOTANICALS”, “A BRIGHTER KIND OF TOMORROW”, “REAL INGREDIENTS. BRIGHTER DAYS.”2. Minimal · gpt-image-2.5-sunburst · high · 1024x1280 · job J34Result: Every requested string is correct. Not asked for: “BOTANICAL SPARKLING WATER”, “REAL BOTANICALS REAL REFRESHMENT”, “BRIGHTER DAYS AHEAD”.3. Bold Gen Z · gpt-image-2.5-sunburst · high · 864x1536 · job J35Result: The brand came out as SUNWELL, because the third prompt never names AURELIS. When you split a post into calls, repeat the shared facts in each. Headline, tagline, icons, 3 flavours and the call to action are correct. Not asked for: “GOOD DRINKS BRIGHTER PEOPLE”, “MORE FLAVOR A BRIGHTER YOU”, “REAL FRUIT. BRIGHTER DAYS.”
Works because. The post holds three separate prompts for one fictional brand (AURELIS): editorial serif, Apple-style minimal, bold Gen Z. Product, icon row, flavour lineup and call to action stay fixed; headline, type style and energy change. That is the fixed/variable split that works for campaign series.
To use. Send each version as its own call. Pasted as one prompt, it asks for three ads in one image.
Watch out. “Aspect ratio: vertical (4:5 or 9:16)” leaves the choice open; set size yourself. The emoji headings are just text.
Sent for the image: The post holds three prompts. Each was sent verbatim as its own call, cut at the version headings (“🌿 2. Minimal Premium” and “🍊 3. Bold Gen Z”), so J33 also carries the post's opening lines.
Prompt, verbatim · 3,368 characters · 478 words
Show the full prose prompt
Create a high-fashion editorial beverage campaign set on a minimalist rooftop at golden hour. Ultra-clean composition with strong negative space, warm sun flare, and soft shadows.
A stylish male model with sharp features and effortless confidence is posed like a fashion editorial—slightly turned away, sipping from the can. He wears elevated linen tailoring in monochrome cream tones, partially unbuttoned, with subtle jewelry.
Foreground: a sculptural stone pedestal holding a large hero can of AURELIS – Blood Orange Basil, hyper-detailed with condensation and botanical artwork.
Typography is bold, oversized, and editorial: GOOD DAYS IN EVERY CAN
Use refined serif typography with dramatic scale and spacing. Keep layout asymmetrical and magazine-like.
Add subtle microcopy: “Botanical. Bright. Made to sparkle.”
Minimal icons (thin line style): LOW SUGAR / REAL FRUIT / LIGHTLY SPARKLING / 12 OZ
Bottom strip: 3 cans styled like a luxury product lineup with fruit + herbs arranged like still life photography:
Blood Orange Basil
Peach Ginger
Lime Mint
CTA: SHOP FLAVORS →
Style: Vogue campaign, luxury branding, cinematic lighting, ultra-realistic, muted tones, high contrast, premium finish.
Aspect ratio: vertical (4:5 or 9:16)
🌿 2. Minimal Premium (Apple-style)
Prompt:
Create a minimalist premium beverage ad with a sunlit rooftop setting and extremely clean composition.
Background: soft gradient sky with warm golden-hour tones, minimal distractions, large negative space.
A relaxed male model sits casually, taking a sip from the can. Wardrobe: simple, modern, neutral linen outfit.
Centerpiece: a large floating or pedestal-mounted can of AURELIS – Blood Orange Basil, perfectly lit with soft reflections and visible condensation.
Typography: SUN IN EVERY SIP
Use modern minimal typography, centered, with generous spacing.
Supporting line: “Real botanicals. Effortless refreshment.”
Icons aligned horizontally: LOW SUGAR • REAL FRUIT • LIGHTLY SPARKLING • 12 OZ
Bottom section: 3 cans evenly spaced, perfectly aligned with subtle shadows:
Blood Orange Basil
Peach Ginger
Lime Mint
CTA button (pill-shaped): SHOP FLAVORS
Style: Apple-level minimalism, ultra-clean layout, soft gradients, modern luxury, highly polished product rendering.
Aspect ratio: vertical (4:5 or 9:16)
🍊 3. Bold Gen Z Premium Version
Prompt:
Create a bold, scroll-stopping vertical beverage ad set on a rooftop during golden hour with vibrant citrus energy.
Lighting: warm, saturated sunlight, glowing highlights, high contrast.
A confident male model drinks directly from the can mid-action, candid and dynamic.
Hero can: oversized, slightly exaggerated scale on a textured stone pedestal, covered in condensation.
Typography: DRINK THE SUN
Use large, expressive typography with strong contrast (deep green + burnt orange), slightly overlapping layout.
Tagline: “Fresh. Bright. Addictive.”
Feature icons arranged dynamically: LOW SUGAR
REAL FRUIT
LIGHTLY SPARKLING
12 OZ
Bottom: 3 cans with energetic ingredient styling:
Blood Orange Basil (splashes, citrus slices)
Peach Ginger (fresh peach + ginger root)
Lime Mint (lime wedges + mint leaves)
CTA: SHOP FLAVORS →
Style: premium but bold, modern campaign aesthetic, slightly experimental layout, high contrast, social-first, dynamic composition.
Aspect ratio: vertical (9:16 preferred)
11
Product lock plus a contrast rule
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1024x1280 · job J36 · input: J27Result: The ring is kept exactly, including the three side diamonds per side from the input. The contrast rule was not followed: a green stone on dark green velvet.
Works because. The first paragraph is a product lock: design, shape, proportions, gemstone arrangement, metal tone, setting. The colour rule is concrete: place the piece against its opposite tone.
Does nothing. “For every generation, create a different visual concept.” Each API call is independent and has no memory of earlier outputs, so it cannot avoid repeating them. Vary the concept yourself: write 4-6 scene lines and send one per call.
Trim. “extraordinary”, “visually unforgettable”, “far beyond ordinary product photography” are praise. The list of supporting props (branch, hand, leaf, stone, silk fold, water ripple) is the useful part.
Sent for the image: Verbatim, with my ring packshot (J27) as the uploaded image.
Prompt, verbatim · 2,880 characters · 413 words
Show the full prose prompt
Use the uploaded image as the one and only product reference. Preserve the jewelry exactly as it is, with high fidelity to its original design, shape, proportions, gemstone arrangement, metal tone, craftsmanship, setting, texture, and identity. Do not redesign, simplify, or alter the jewelry itself in any way. Keep the product accurate, luxurious, and instantly recognizable.
Create an extraordinary luxury jewelry campaign image where the product is the absolute visual hero. Build a bold, artistic, and premium scene around it that feels cinematic, elegant, and visually unforgettable. The result must never feel like a basic catalog shot or a repetitive product render.
For every generation, create a different visual concept so the outputs do not look similar to one another. Vary the composition, environment, supporting element, texture, background structure, framing, angle, and styling approach each time. Each image should feel unique, fresh, and creatively elevated while still maintaining a refined luxury identity.
Include one or more strong supporting natural or tactile elements that help frame and enhance the jewelry, such as a branch, hand, leaf, stone, bark, flower petal, sand texture, silk fold, glass reflection, water ripple, smoke, shell, or sculptural organic form. These elements should not distract from the product, but should artistically support it and make it feel more premium, emotional, and visually magnetic.
Use color contrast intelligently. Place the jewelry within a scene that uses an opposite or contrasting color tone to make the piece stand out strongly, while still keeping the palette harmonious, tasteful, and luxurious. The contrast should feel intentional and sophisticated, never random or harsh. The product must pop clearly from the scene through contrast in color, texture, light, or material.
Use strong visual hierarchy, elegant negative space, and a striking focal composition that makes the jewelry dominate the frame. The product should feel iconic, powerful, and highly desirable. Emphasize macro-level detail, realistic sparkle, gemstone brilliance, polished metal reflections, fine craftsmanship, prongs, edges, texture, and premium material depth.
Lighting should be cinematic and refined, with soft directional light, controlled highlights, elegant shadows, subtle rim light, atmospheric glow, and beautiful depth. Use shallow depth of field and macro product-photography aesthetics to keep the jewelry crisp and visually commanding.
The final image should feel like a world-class luxury editorial ad from a top creative studio: visually bold, highly refined, emotionally captivating, and far beyond ordinary product photography.
Avoid repeated concepts, repeated props, repeated backgrounds, flat lighting, weak framing, visual clutter, cheap styling, generic catalog presentation, text, watermark, and logos.
gpt-image-2.5-flare · medium · 1024x1536 · job J37Result: 1 citrus wedge, 1 mint cluster, 1 amber bracelet, and the hand enters from the lower left gripping the stem. It drew 3-4 bulbs instead of 2.
Works because. Every prop that could multiply has a number: “1 visible citrus wedge”, “1 fresh mint garnish cluster”, “1 chunky translucent amber bracelet”, “2 warm exposed hanging bulbs”.
Carries the effect. The hand has an entry point and a grip: “enters from the lower left, delicately gripping the stem”. Hands go wrong less often when the pose is specified.
To reuse. Swap the drink and the location. Keep the counts, the backlight sentence and “shot from a low angle looking upward at the raised glass”.
Prompt, verbatim · 1,226 characters · 193 words
Show the full prose prompt
A cinematic vertical photo of a hand holding up a large balloon wine glass filled with a sparkling golden-yellow citrus cocktail in a lush indoor greenhouse bar. The drink is backlit by warm late-afternoon sunlight, making it glow translucent amber. Inside the glass there is 1 visible citrus wedge, and at the rim there is 1 fresh mint garnish cluster. The hand enters from the lower left, delicately gripping the stem, wearing 1 chunky translucent amber bracelet. The setting is dense with tropical greenery, hanging ferns, and vine-covered walls, with a bright greenhouse roof structure visible overhead and 2 warm exposed hanging bulbs softly glowing in the background. Use shallow depth of field with creamy bokeh, strong sun rays filtering through leaves, soft haze, and rich green-and-gold color contrast. Add a blurred foreground leaf or plant along the right edge to frame the composition. The lower background should suggest a busy café or cocktail lounge with indistinct people, but keep them heavily out of focus. Photorealistic, elegant lifestyle photography, moody yet sun-drenched, shot from a low angle looking upward at the raised glass, high detail on condensation, glass reflections, and the luminous drink.
gpt-image-2.5-flare · medium · 1024x1024 · job J38Result: The feet are swapped: the loafer is on her left foot and the right foot is bare. The spare loafer lies lower left with the mule heel and horsebit.
Works because. It says which foot wears what: “One loafer is worn on the right foot and the left foot is bare ... while the second matching loafer lies on the rug in the lower left foreground”. Without this, models put shoes on both feet or draw two pairs.
Also counts. “1 pair”, “2 thin irregular black lines” on the rug.
To reuse. A template for any worn product (shoes, watches, bags): which body part wears it, where the spare sits, and the product in construction terms (almond toe, open mule heel, horsebit hardware).
Prompt, verbatim · 1,148 characters · 191 words
Show the full prose prompt
A warm, editorial-style lifestyle product photo shot indoors from a low close-up angle, focused on a woman's lower legs and feet as she tries on 1 pair of black leather backless loafers with tan faux-fur lining. One loafer is worn on the right foot and the left foot is bare, hovering just above the textured cream shag rug, while the second matching loafer lies on the rug in the lower left foreground. The shoes have smooth black leather uppers, a rounded almond toe, open mule-style heel, plush brown fur spilling out around the opening, and a small polished gold horsebit hardware detail across the vamp. The model wears cropped medium-blue denim jeans with a raw frayed hem. The setting is a cozy minimalist interior with a cream rug featuring 2 thin irregular black lines, a neutral wall, and a leaning rectangular mirror with a medium wood frame in the upper right background, softly reflecting the rug and part of the scene. Use soft natural window light, shallow depth of field, subtle film grain, realistic skin texture, muted beige and black palette, relaxed candid composition, premium fashion catalog mood, high detail, photorealistic.
Typography posters
Six text-heavy layouts in English, Spanish, Japanese and Traditional Chinese. All of them quote every string, give it a position, and count the elements.
YouArt's own example image for this category, on GitHub (not copied here): poster.webp
gpt-image-2.5-sunburst · high · 1024x1536 · job J10Result: Every requested string is there, the 6 product cards are named as written and the face is covered. Not asked for: four more handwritten notes around the layout.
Works because. Every text block has a position and a count, and the actual Japanese strings are in the JSON. With 6 product cards, a note, a concept box and a badge, the structure keeps each string attached to its place.
Carries the effect. The count fields. Counting is how you stop the model from adding a seventh product card.
Privacy device. The character's face is “intentionally obscured with a plain beige rectangle”. Several prompts in the repo use this.
Check. Small Japanese text is the weak point. Read every string at 100% zoom, and use high quality for small text.
Not valid JSON. As published, the prompt is missing its final closing brace. The image model does not mind; a JSON validator in a script pipeline would reject it.
Sent for the image: Template slots filled with their defaults.
gpt-image-2.5-sunburst · high · 1024x1280 · job J11Result: All 6 panel elements as written, including the Japanese; neon script title; two large palms plus small ones.
Works because. Goal, Canvas, Main image, Background, Typography, Bottom panel, Visual style, Constraints. Labeled sections give the same structure as JSON without the braces.
Carries the effect. “exactly 6 visible text/icon elements”, numbered 1-6, each with colour and content, and “exactly 2 prominent palm tree silhouettes”. The Constraints section repeats the rules: no extra logos, keep the Japanese exactly.
To reuse. A good skeleton for any cover or thumbnail. Compare it with the jellyfish JSON: same kind of job, other format.
Sent for the image: Template slots filled with their defaults.
Goal: Create a square-to-portrait neon music cover poster for a tropical synthwave dance track titled {argument name="headline text" default="Miami Night"}, blending a silhouetted dancing woman with a Miami skyline double exposure.
Canvas: Vertical 4:5 poster, high-resolution, glossy cinematic finish. Use a vivid sunset-to-night gradient background: hot orange and magenta near the horizon, deep purple and navy starry sky at the top.
Main image: Center a full-body female dancer silhouette, facing right in profile, arms raised above her head with both hands posed like she is dancing. Her long loose hair flows left in the wind. Her body is a dark silhouette filled with a double-exposure city scene: illuminated high-rise skyscrapers, tiny window lights, pink rooftop lights, palm trees, waterfront reflections, and rippling bay water inside the torso and legs. Keep her outline crisp and elegant, with strong backlighting.
Background environment: Show a Miami waterfront skyline at night with palm trees on both sides, blurred city buildings, sparkling lights, and neon reflections across the water. Include exactly 2 prominent palm tree silhouettes in the foreground, one on the left and one on the right, plus smaller distant palms near the skyline. Add a star-filled sky above.
Typography over main art: Across the dancer and skyline, place large glowing handwritten cursive neon text reading {argument name="neon script text" default="Miami Night"}. The script should be bright pink-white, like a neon tube, spanning nearly the full width of the image and crossing the center.
Bottom information panel: Add a dark navy rectangular lower panel occupying about the bottom quarter of the poster. Include exactly 6 visible text/icon elements arranged like a music promotion card: 1) a pink music note icon followed by white title text {argument name="track title" default="MiamiNight"}; 2) pink subtitle text {argument name="challenge text" default="“Dance party” Challenge"}; 3) a thin decorative horizontal divider with a small sparkle at the center; 4) white artist line reading {argument name="artist line" default="— Crystal Dragon Lancer エル"}; 5) two pink hashtags reading "#Suno" and "#AI音楽"; 6) small white credit text at bottom right reading "Direction / Prompt: Elle".
Visual style: Retro Miami nightlife, synthwave album artwork, neon pink and purple color palette, cinematic double exposure, glossy reflections, high contrast, dreamy tropical party atmosphere, clean social-media cover design.
Constraints: Keep all listed text legible and positioned as described. Do not add extra logos, watermarks, or unrelated text. Preserve the Japanese characters exactly where they appear in the default text.
gpt-image-2.5-sunburst · high · 1024x1536 · job J39Result: All requested strings, 3 signs, 4 awards, 5 stars, the “..” copied as written, and the beige face block. Not asked for: Japanese text, “ALGUNOS LUGARES SIEMPRE VUELVEN”, “UN VIAJE. INFINITAS VERSIONES DE TI.” and invented credit names.
Works because. Every text element is quoted and placed: tagline top center, awards upper left, title lower center, quote beneath it, release line at the bottom.
Carries the effect. “exactly 3 directional signs” and “exactly 4 award blocks”, each with its exact strings. Counts plus strings mean fewer invented laurels.
Watch out. One paragraph with about 12 strings is hard to edit. Split it into labeled sections before you change anything.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 2,738 characters · 428 words
Show the full prose prompt
A cinematic anime movie poster for a fictional film titled {argument name="headline text" default="EL VIAJE DE LA LUNA DE PLATA"}, in polished modern Japanese animation style with a natural, less over-detailed look. Center a teenage anime girl from mid-thigh up, facing forward, with a short silver bob haircut, pale skin, a black choker, small black geometric earrings, a white tank top, and a dark navy oversized zip hoodie with two yellow stripes running down the sleeves. She has a backpack strap over one shoulder and both hands tucked casually into the hoodie pockets. Her face is obscured by a flat rectangular censor block in a muted beige tone, covering the entire face area. Place her in a dramatic twilight coastal city setting that blends travel, nostalgia, and fantasy: on the left, a lit train platform with a commuter train approaching, its destination sign showing Japanese characters; behind it, a glowing city skyline with a ferris wheel. In the distance and lower left, layered mountains and a winding illuminated valley road. On the right, a cliffside coast at sunset with the sea reflecting warm light, a crescent moon in the sky, several flying seabirds, and a curving highway descending along the hillside. Also on the right, include a wooden signpost with exactly 3 directional signs labeled "NUEVOS CAMINOS", "VIEJOS RECUERDOS", and "SIN LÍMITES". At the top center, add the Spanish tagline {argument name="tagline text" default="CADA DESTINO CAMBIA SU HISTORIA"} in elegant serif capitals. On the upper left, create an awards column in gold typography with laurel wreaths and exactly 4 award blocks: one text block reading "GANADORA DE MÚLTIPLES PREMIOS" with 5 gold stars beneath it, then three laurel award sections reading "MEJOR PELÍCULA ANIMADA / FESTIVAL INTERNACIONAL DE ANIMACIÓN / 2024", "PREMIO DEL PÚBLICO / FESTIVAL INTERNACIONAL DE CINE / 2024", and "MEJOR BANDA SONORA ORIGINAL / ACADEMIA DE CINE ANIMADO / 2024". Place the film title large across the lower center in luminous ornate serif lettering with a magical glow and sweeping flourishes, layered partly over the character. Beneath it, add the Spanish quote {argument name="quote" default="A veces, para encontrarte... tienes que perderte en el mundo."}. Below that, add "UNA PELÍCULA DE ESTUDIO LUMINARIA" in small caps. At the bottom, add the release line {argument name="release text" default="PRÓXIMAMENTE EN CINES"} in large gold serif capitals, plus tiny production logos and credits along the footer, including a small studio emblem on the left. Rich blue, violet, and warm sunset orange palette, glossy poster lighting, romantic adventure mood, balanced composition, highly polished theatrical key art, vertical one-sheet film poster.
gpt-image-2.5-sunburst · high · 1536x1024 · job J40Result: 9 panels spelling MADPENCIL. The N (a braided river) is hard to read. The hashtags were not rendered. Wide white margins.
Works because. One line per panel: letter, biome, and how the terrain forms the letter. The repeated sentence shape keeps 9 panels consistent.
Carries the effect. “no text overlays” together with “letters legible yet organically merged with the landscape”. The letter has to come from the land, not from a font laid on top.
To reuse. Change the word and pick one biome per letter. Nine panels is already a lot. Drop “ultra high resolution 8K” and the hashtags at the end; both are read as text.
Prompt, verbatim · 1,752 characters · 261 words
Show the full prose prompt
Ultra-realistic overhead satellite view from orbit, a crisp modern editorial layout featuring 9 vertical panels arranged side by side on a white background, together spelling "MADPENCIL", each panel containing a single letter formed entirely from real Earth terrain and natural topography, no text overlays:
Panel 1 (M): jagged mountain ridges and deep ravines composing a sharp angular "M", dramatic shadows, rocky surface
Panel 2 (A): a meandering river through thick tropical forest shaping an "A", vivid contrast between water and canopy
Panel 3 (D): vast desert sand dunes sculpted by wind into a smooth "D", warm earthy tones, gentle gradients
Panel 4 (P): patchwork cropland and geometric farm fields arranged into a structured "P", clean grid patterns
Panel 5 (E): glacial formations and ice sheets carving a crisp "E", pure whites and deep icy blues, cracked textures
Panel 6 (N): a braided floodplain river system forming "N", branching channels and natural flow patterns
Panel 7 (C): a curved coastline and ocean meeting point shaping "C", wave breaks and sediment visible
Panel 8 (I): a narrow slot canyon or straight waterway forming a minimal "I", strong vertical line
Panel 9 (L): volcanic landscape with hardened lava flows forming an "L", dark basalt with glowing lava accents
straight-down satellite perspective, NASA Earth imagery aesthetic, hyper-detailed terrain, realistic geography, consistent lighting and scale across panels, minimal cloud cover, high contrast, sharp resolution, subtle atmospheric haze, true-color grading, ultra high resolution 8K, clean panel spacing, gallery-style composition, visually unified but each panel a distinct biome, letters legible yet organically merged with the landscape
#AIart #GPTImage2
gpt-image-2.5-sunburst · high · 1024x1536 · job J41Result: All requested Chinese strings, 5 stops with times, markers in the same order, 2 stats and a compass. Not asked for: “雲海之上 遇見更美的世界”, “阿里山·遇見更美的台灣”, “小笠原山 海拔 2,488公尺”, “阿里山·心之所向”, “阿里山神木” and a train number.
Works because. The same 5 stop names appear twice, on the itinerary card and on the map markers, and the prompt says the markers must match the list in order.
Carries the effect. Each stop has a name, a time and a description in 「」 quotes, plus “exactly 5” stops and “exactly 2 stats”.
To reuse. Swap the destination and the stops. With more than 20 Chinese strings, proofread each one in the output.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 2,878 characters · 433 words
Show the full prose prompt
Create a vintage illustrated travel poster in traditional Chinese for {argument name="destination name" default="阿里山國家風景區"}, designed as a one-day itinerary infographic with a split vertical layout. The left panel is a parchment-textured itinerary card in warm beige with ornate gold Art Nouveau borders and dark brown typography, and the right panel is a dramatic painted fantasy-realism map scene of a mountain journey at sunrise and sunset tones. At the top of the left panel, large headline text reads {argument name="headline text" default="阿里山國家風景區一日遊"}. Beneath it, include a short centered tagline in traditional Chinese: 「一座高山,五個經典景點。難忘的奇幻旅程。」 with a small decorative mountain divider. The left panel must contain exactly 5 numbered itinerary stops stacked vertically, each with a circular black-and-gold number badge, a small vignette illustration, a bold location name, a time in parentheses, and a short Chinese description. The 5 stops are: 1. 「阿里山車站」 at 「(8:00 AM)」 with a wooden mountain railway station illustration and description 「開啟探索神木與森林的旅程。」 2. 「阿里山森林鐵路」 at 「(9:30 AM)」 with a red-and-black steam train illustration and description 「穿越森林,體驗百年林鐵風情。」 3. 「神木區棧道」 at 「(11:30 AM)」 with giant cedar trees and elevated wooden boardwalk illustration and description 「漫步千年巨木下,感受森林靈氣。」 4. 「姊妹潭」 at 「(1:30 PM)」 with a tranquil forest lake and pavilion illustration and description 「欣賞靜謐湖光,聆聽自然樂章。」 5. 「小笠原山展望台」 at 「(4:00 PM)」 with a wooden observation deck above clouds at sunset illustration and description 「觀賞壯闊山景與雲海,欣賞日落。」 The right panel should depict a continuous glowing golden path winding through exactly 5 numbered map markers that match the left panel labels in order, with black-and-gold marker plaques reading: 1 「阿里山車站」, 2 「阿里山森林鐵路」, 3 「神木區棧道」, 4 「姊妹潭」, 5 「小笠原山展望台」. Show stop 1 as a rustic alpine wooden station perched on a cliff among pine forests; stop 2 as a small steam locomotive traveling on a curved mountain railway with smoke drifting upward; stop 3 as towering ancient red cypress trees with a spiral and zigzag wooden walkway around the trunks; stop 4 as an emerald lake surrounded by dense forest with a small pavilion and arched bridge; stop 5 as a lookout deck on a peak above a sea of clouds, facing a glowing sunset. The environment should feature layered mountain ranges, mist-filled valleys, evergreen forests, golden-hour light, luminous cloud seas, and a romantic painterly atmosphere with rich detail. At the bottom right, add a decorative compass rose labeled N, E, S, W, plus a dark green and gold information box with exactly 2 stats in traditional Chinese: 「總距離 ~9公里 / 5.6英里」 and 「預計時間 全天 - 14,500步」. Overall style: premium tourism poster, painterly digital illustration, nostalgic national-park brochure aesthetic, highly detailed, warm sepia and gold accents, elegant composition, readable Chinese text, vertical 2:3 poster.
gpt-image-2.5-sunburst · high · 1536x864 · job J42Result: The literal \n became a line break. All four blocks are correct, with the coral highlight on 具体的な and one check icon. The coral underline sits under “の次は？”. The woman is fictional.
Works because. The Constraints line counts everything: one person, four text blocks, one check icon, three accent types. The layout is given as a split (40% portrait, 60% text).
Carries the effect. Font per block: Mincho-style serif for the headline, sans-serif for the labels, and which phrase gets the coral highlight.
Watch out. The headline contains a literal \n. Say “on two lines, break after 基準画像を” instead.
Sent for the image: Verbatim, including the literal \n in the headline, to see what the model does with it.
Prompt, verbatim · 2,244 characters · 341 words
Show the full prose prompt
Goal: Create a clean Japanese promotional thumbnail/banner for a GPT Image 2 tutorial about using reference images to keep the same person consistent.
Canvas: 16:9 horizontal banner, 1200×675 style, soft pale blue and white background with large organic blob shapes. Use a polished educational course-thumbnail aesthetic, airy spacing, high readability, no harsh shadows.
Layout: Split composition with a realistic portrait on the left occupying about 40% of the canvas, and a text-heavy information panel on the right occupying about 60%. The right side has a large white rounded organic shape as the main text area. Add small decorative dotted grids in the top-right and bottom-left corners, a light diagonal hatch pattern near the lower-right edge, and coral accent marks beside the headline.
Subject details: On the left, show one smiling Japanese woman in her late 20s to early 30s, photographed from the waist/chest up, facing the camera. She has shoulder-length softly layered black hair, natural makeup, warm friendly expression, and wears a peach sheer button-up shirt over a white ribbed camisole with a delicate gold necklace. Background behind her is bright, soft-focus, and slightly green-white like daylight through a window.
Text content: Include exactly 4 text blocks on the right: 1) a dark navy rounded pill at the top reading 「GPT Image 2 実践編 ③」; 2) a large navy Japanese headline reading 「基準画像を\n作った、その次は？」 with a coral underline under the final phrase; 3) a thin outlined rectangle subtitle reading 「人物固定の具体的な使い方」 with 「具体的な」 highlighted in coral; 4) a bottom checklist row with a pale blue circular check icon followed by 「毎回どの基準画像を使う？」. Keep all Japanese text crisp, centered within its area, and use elegant Mincho-style serif for the main headline with modern sans-serif for smaller labels.
Color palette: Deep navy text, pale powder blue background, white panels, peach/coral accents matching the woman's shirt, and light blue-gray icon details.
Constraints: Use exactly one person, exactly four text blocks, exactly one check icon, and exactly three main decorative accent types: dotted grids, coral rays/underline, and diagonal hatching. Avoid extra logos, watermarks, English translations, or additional text.
Cinematic scenes
Five single frames. They get the film look from named light sources and lens terms, not from the word “cinematic”.
YouArt's own example image for this category, on GitHub (not copied here): travel.webp
gpt-image-2.5-sunburst · high · 1536x864 · job J12Result: Close to the prompt. Not asked for: invented signage and captions (“SECTOR A-7”, “KEEP CLEAR”, “LAST HORIZON”, “THE OCEAN REMEMBERS”).
Works because. Comma-separated blocks in a fixed order: character, pose, world, light, lens, style, separated by “ , ”. Dense, but each block has one job.
Carries the effect. Named light sources with colours: cold teal ambient from the sky, warm amber sodium glow camera-right, hard backlight from a low sun. Lens terms (“35mm anamorphic”, “horizontal lens flares”, “slight low angle”) set the film look.
To reuse. Keep the light and lens blocks, replace the character and the world. “Format 16:9” is text; set the size.
Prompt, verbatim · 2,202 characters · 338 words
Show the full prose prompt
A mecha girl mid-teens, pale skin smudged with soot and salt spray, sharp amber eyes with glowing HUD reticles, waist-length ash-white hair tied in a high ponytail whipping in the sea wind, matte gunmetal exoskeleton armor plating her shoulders, forearms and shins, exposed hydraulic pistons at the joints, chest rig with glowing cyan coolant lines, oversized oil-stained hangar jacket half slipping off one shoulder, a massive rail cannon resting on her right shoulder, dog tags and frayed red ribbon at her collar , standing off-center to the left on the rusted edge of a tilted steel platform jutting out over dark water, weight shifted onto one leg, left hand gripping the cannon strap, head turned slightly toward camera with a quiet defiant stare, steam venting from her back thrusters, her ponytail and jacket streaming sideways in the salt wind , a vast derelict sea-city at dusk, colossal megastructures of unknown purpose rising from the ocean in staggered silhouettes, bone-white monolithic towers fused with barnacled steel, cyclopean ring-shaped constructs canted at broken angles, rusted skeletal gantries threaded with dead cables, dark swells rolling between the pylons, shipwrecks half-swallowed at their feet, thick sea fog clinging to the bases while the upper structures pierce into a bruised sky, scattered faint lights blinking high in the towers like distant eyes , moody low-key lighting, cold teal ambient from the overcast sky, warm amber sodium glow leaking from a distant structure camera-right, hard backlight from a low sun behind the towers carving her silhouette, volumetric god rays cutting through sea mist, wet specular highlights on her armor , 35mm anamorphic lens, slight low angle looking up past her shoulder toward the structures, medium-wide shot, shallow depth of field with foreground rust in soft focus, horizontal lens flares, fine atmospheric haze compressing the distant megastructures into layered silhouettes , cinematic anime key visual, painterly digital illustration with crisp line art, desaturated oceanic palette of teal, bone-white and rust punched by small warm accent lights, film grain, high-contrast editorial poster aesthetic . Format 16:9.
gpt-image-2.5-sunburst · high · 1536x1024 · job J13Result: Second attempt. The first put a real space-agency logo on the suit; this one has a flag patch.
Works because. Layout first (high-angle full shot, fire at the exact center between two figures), then the left figure, the right figure, the central effect and the light.
Carries the effect. The one surreal element is described as a physical transition: “the warm orange flames should transition seamlessly into a photorealistic spiral galaxy”.
Weak part. The last paragraph lists themes (human connection, cosmic wonder). A theme is not something the model can draw; it adds little.
Used again. This image is the input for the contact sheet card below (J15).
Prompt, verbatim · 1,619 characters · 257 words
Show the full prose prompt
Create a surreal, cinematic, photorealistic 8k conceptual artwork set in an abstract cosmic void. Use a high-angle full shot with symmetrical balance, placing a magical campfire at the exact center between two seated characters against a pitch-black background with no visible environment.
On the left, show an astronaut sitting on a rough gray rock, wearing a highly detailed realistic white space suit with the helmet on and a reflective visor. The astronaut holds a stick and roasts a marshmallow over the fire.
On the right, show a young man sitting on a metal folding chair and leaning forward toward the flames. He has dark hair and wears a red and black plaid flannel shirt, blue jeans, and brown boots. He is warming his hands by the fire.
Make the central campfire magical and surreal: the warm orange flames should transition seamlessly into a photorealistic spiral galaxy filled with glittering stars, cosmic dust, and visible swirling motion. The galaxy should rise and arch overhead like a celestial canopy.
Lighting should be dramatic and cinematic. Use warm orange firelight as the primary light source, with cool blue and purple glow from the galaxy as secondary light. Create soft but deep shadows and strong contrast. Use a color palette dominated by black, orange, blue, and purple, with white, red, and brown accents.
The image should communicate human connection, cosmic wonder, the contrast between ordinary life and the infinite universe, and shared warmth in an endless dark space. Render with ultra-detailed photorealistic textures, cinematic lighting, sharp focus, and high dynamic range.
gpt-image-2.5-sunburst · high · 1536x1024 · job J43Result: Second attempt; the first had a real beer brand's sign on a building. The police markings, Skytree and the Asakusa, Kuramae, Komagata Bridge and Sumida River signs follow from the defaults, which name the real Tokyo police. 5 workers and about 7 debris types. Not asked for: invented slogans on a banner and a sign.
Works because. It treats a fantasy subject as a news photo: telephoto, street level, rain streaks, emergency lights reflecting on wet surfaces. The realism comes from the capture, not from the robot.
Carries the effect. Counts with positions: “exactly 7 distinguishable item types” in the debris, “exactly 5 human emergency workers”, with who stands where.
To reuse. The slots at the end (robot, action, location, weather and mood) are the variable fields; the capture sentences stay.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 2,386 characters · 338 words
Show the full prose prompt
Create a cinematic photorealistic disaster-response scene in rainy Tokyo: a massive bipedal police construction robot from the {argument name="organization" default="Tokyo Metropolitan Police Department heavy machinery countermeasure robot unit"} stands waist-deep in a flooded urban river or canal, clearing a huge debris jam that is damming the water. The robot is seen from behind-left at street level, towering over the scene, with bulky white-and-black industrial armor plates, exposed hydraulics, pistons, cables, tracked-looking leg components, mud splatter, red emergency beacons glowing on both shoulders, and visible markings reading 「警視庁」 and “MPD” on its back armor. It grips and lifts one enormous tangled debris pile containing exactly 7 distinguishable item types: broken tree branches and logs, bamboo-like sticks, dead leaves and vines, bent metal railings, a blue plastic crate, a white rectangular signboard, and a bicycle wheel/frame. The floodwater is brown, violent, and foamy, splashing around the robot’s legs and pouring from the lifted debris. In the foreground, show exactly 5 human emergency workers/police observers in helmets and rain gear behind metal barricades: 2 navy MPD officers with white helmets at lower left, 1 pointing toward the robot, 1 orange-and-navy rescue worker at lower right, and 2 partially visible helmeted workers near the bottom edge. Add a crowd of civilians with umbrellas along the left sidewalk behind barriers. The city background should feature tall office buildings, wet reflective streets, heavy rain streaks, mist, evening-gray light, and urban signage including a vertical Japanese police slogan sign on the left and a blue road sign on the right indicating Asakusa, Kuramae, and route 6, plus a bridge or wall label for a river. Use a realistic telephoto news-photo look, high detail, dramatic scale, shallow atmospheric depth, water droplets on surfaces, emergency lights reflecting in rain, no cartoon styling. Main customizable subject: {argument name="robot description" default="a massive white-and-black MPD heavy rescue mech"}. Main action: {argument name="action" default="clearing a debris dam from a flooded river"}. Location: {argument name="location" default="rainy Asakusa/Kuramae, Tokyo"}. Weather and mood: {argument name="weather and mood" default="torrential rain, chaotic but heroic disaster response"}.
gpt-image-2.5-flare · medium · 864x1536 · job J44Result: Every count met: 3 panels with dividers, 2 whales, 1 woman, 1 astronaut, 1 ringed planet, 4 moons, no text.
Works because. “exactly 3 equal-width vertical scenes”, then one section per panel with its own palette, and one shared style section.
Carries the effect. The Constraints section limits people to one woman in the center and one astronaut on the right. Without it, fantasy scenes fill up with extra figures.
To reuse. Written for a 9:16 phone wallpaper. For a wide triptych, the API allows up to 3:1.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 2,769 characters · 409 words
Show the full prose prompt
Goal: Create an ultra-detailed fantasy triptych wallpaper made of exactly 3 tall vertical panels placed side by side with thin black dividers, each showing a different breathtaking world in cinematic realism.
Canvas: Tall portrait aspect ratio, about 9:16, high-resolution, immersive, sharp foreground and deep atmospheric perspective. No text, no logo, no watermark.
Layout: Divide the image into exactly 3 equal-width vertical scenes: left underwater, center mountain fantasy, right alien space landscape. Each panel should feel like its own epic environment while sharing the same luminous, magical, highly polished digital-art style.
Left panel — underwater civilization: Show a glowing blue ocean canyon with shafts of sunlight streaming down from the surface. Include exactly 2 large whales swimming near the top and mid-height, multiple translucent jellyfish of varying sizes, many small tropical fish, coral reefs, rocky arches, and a majestic submerged palace city with domes, towers, bridges, and warm lights. The color palette is turquoise, cyan, deep blue, and bioluminescent white.
Center panel — cherry blossom immortal mountain realm: Show an ancient East Asian pavilion balcony on a cliff above a sea of clouds at sunset. A lone robed woman with very long dark hair sits or kneels on the cliff edge, looking toward distant misty karst mountains and waterfalls. Include exactly 1 main character, exactly 1 large cherry blossom tree arching across the top with pink blossoms, drifting petals, ornate wooden roof eaves, lanterns, distant temples, and golden-pink sunset light. Use the mood of {argument name="fantasy mood" default="peaceful celestial wonder"}.
Right panel — alien ringed planet frontier: Show a snowy extraterrestrial mountain range under a star-filled sky with a giant ringed planet dominating the upper right. Include exactly 1 astronaut standing on a rocky snowy cliff in the foreground, looking out over an icy valley and glowing horizon. Include exactly 1 giant ringed planet, exactly 4 smaller moons or planets, a bright Milky Way band, and a few shooting stars. The palette is icy blue, violet, white, and cosmic indigo.
Visual style: Epic fantasy sci-fi concept art, photorealistic detail blended with painterly grandeur, dramatic lighting, volumetric rays, glowing highlights, crisp textures, high dynamic range, vertical poster composition, awe-inspiring scale. Emphasize {argument name="overall style" default="cinematic hyper-detailed digital painting"} and a unified dreamlike atmosphere.
Constraints: Use exactly 3 panels; keep the thin black vertical borders visible; do not add captions or typography; avoid extra human figures beyond the single woman in the center panel and the single astronaut in the right panel.
gpt-image-2.5-flare · medium · 1536x1024 · job J45Result: Girl with a cup on the left, two puppies in the river, the four named colours and bold ink lines. The background is not blurred, so the shallow depth of field was ignored.
What it says. The first line, translated: “Trying a fixed-pattern GPT Image 2 prompt structure, one close shot, one medium shot.” Then 10 fields: subject (an anime girl holding coffee by a river, watching two puppies play in the water), scene, composition (medium shot, subject left, 50mm, shallow depth of field), light (soft sun from the upper right), style (hand-drawn line illustration), colour, detail, mood, rendering, avoid (text, watermark, logo, malformed structure).
Works because. 272 characters, the shortest prompt here, and still complete. Fixed fields make a short prompt cover everything without JSON.
To reuse. Keep the 10 field names, in any language. Four named colours (peacock blue, mustard yellow, coral pink, ivory) set the palette better than a mood word.
Six multi-panel prompts. The repo has no storyboard category, so these come from character design, illustration and infographic rows. Grids are where structure, JSON or not, pays off most.
YouArt's own example image for this category, on GitHub (not copied here): storyboard.webp
gpt-image-2.5-sunburst · high · 1536x864 · job J14Result: 12 panels in a 3x4 grid, warning text exact. Not asked for: a caption on every panel and extra labels in panel 8.
Works because. A 3x4 grid with one description per cell, and a continuity line saying all 12 panels are one sequence from entry to reveal.
Carries the effect. The vehicle block is stated once and applies to every panel (compact armored ship, 3 bright rear engines). The warning panel counts its text (4 lines + 1 readout) and lists the exact strings.
To reuse. Keep layout.grid, continuity and lighting; rewrite the 12 sections.
Line breaks added for reading. Copy gives the original single line.
{
"type": "cinematic sci-fi storyboard contact sheet",
"subject": {
"primary": "a small futuristic spacecraft descending into a massive gas giant storm system",
"secondary": "an enormous leviathan-like silhouette hidden within the clouds",
"mood": "oppressive, catastrophic, awe-struck, high tension, cosmic dread",
"style": "photorealistic cinematic concept art with dark sci-fi realism, volumetric storm clouds, strong contrast, amber and black palette with occasional cold blue lightning",
"aspect_ratio": "16:9"
},
"vehicle": {
"design": "compact armored deep-atmosphere ship with 3 bright rear engines, angular industrial hull, worn metallic panels",
"scale": "tiny compared to the planet and creature"
},
"layout": {
"grid": {
"rows": 3,
"columns": 4,
"count": 12
},
"sections": [
{
"position": "row 1 col 1",
"description": "wide exterior shot of the ship entering the upper atmosphere of a colossal gas giant at extreme speed, glowing clouds streaked with fire and friction around the vessel, curved planetary horizon visible"
},
{
"position": "row 1 col 2",
"description": "cockpit POV, dark interior filled with red and cyan holographic instruments, forward visibility collapsing into turbulent storm layers and electrical haze"
},
{
"position": "row 1 col 3",
"description": "exterior mid-wide shot of the ship diving into a gigantic rotating cloud funnel, surrounded by violent spiraling storm structure"
},
{
"position": "row 1 col 4",
"description": "extreme close exterior of the ship hull as bright lightning strikes dangerously close, white electric energy crawling across the metal surface"
},
{
"position": "row 2 col 1",
"description": "dashboard warning screen in red, showing a critical systems failure interface with the exact visible text count of 4 warning lines and 1 large percentage readout: ['WARNING','ENGINES COMPROMISED','THRUST FLUCTUATION','GRAVITY SPIKE DETECTED','DESCENT RATE -453%']"
},
{
"position": "row 2 col 2",
"description": "rear three-quarter exterior of the ship fighting turbulence inside dense storm clouds, engines burning hard while the craft barely holds course"
},
{
"position": "row 2 col 3",
"description": "massive circular disturbance forming in the clouds like an eye or maw, entire storm systems displaced by something huge moving beneath"
},
{
"position": "row 2 col 4",
"description": "second cockpit view with radar-like navigation display and red alert text, pilot making a blind evasive maneuver through lightning-filled darkness"
},
{
"position": "row 3 col 1",
"description": "first reveal of the colossal creature shape rising near the ship, black organic surface and immense curved anatomy emerging from darkness, ship tiny at lower left"
},
{
"position": "row 3 col 2",
"description": "spiral descent shot, ship caught inside a vortex tunnel of clouds, spinning downward with engines flaring as it struggles to recover"
},
{
"position": "row 3 col 3",
"description": "sudden breakthrough into a calm void, minimal composition, ship flying in eerie silence through dark open space with soft mist and no visible storm around it"
},
{
"position": "row 3 col 4",
"description": "final reveal, gigantic leviathan fully emerging behind or beside the ship in cleared space, backlit by a pale circular storm opening, enormous open maw-like silhouette dwarfing the craft"
}
],
"continuity": "all 12 panels depict one continuous descent sequence from atmospheric entry to final creature reveal"
},
"lighting": {
"primary": "glowing amber storm light",
"secondary": "red cockpit interface glow",
"accents": "blue-white lightning and engine exhaust"
},
"environment": {
"location": "inside the upper and middle storm layers of a gigantic gas giant",
"weather": "violent turbulence, electrical storms, vortex funnels, cloud walls, pressure chaos",
"threat": "no safe zone, repeated near-failure, unknown colossal presence driving the storm"
}
}
gpt-image-2.5-sunburst · high · 1536x1024 · job J46Result: Title, legend, captions and infographic strings all correct. 8 panels with badges and ~1.5s tags. The cast stays consistent, no real branding, faceless crowd.
Works because.== TITLE BAR ==, == MAIN GRID ==, == STRICT EXCLUSIONS ==: plain-text section headers. Each of the 8 panels gets a caption string, an icon, a shot and the cast.
Carries the effect. CHARACTER NOTES repeats the cast with fixed colours (red-jersey cat, blue-jersey dog, hamster in a yellow kit) and “Each animal identical across panels”. Consistency by repetition.
Exclusions done right. Each item blocks a real failure: no real team logos, no FIFA/UEFA branding, no jersey numbers, no real stadium text. Long, but every line is specific.
A professional pre-production storyboard sheet for a 12-second photorealistic short film, single landscape page, layout top to bottom:
== TITLE BAR ==
Black bar across full width. Left: massive bold white text "PET WORLD CUP". Right: two outlined boxes "TOTAL VIDEO TIME: 12 SECONDS" and "8 SHOTS · ENSEMBLE · COMEDY · CINEMATIC".
== ICON LEGEND ==
Four icon+label pairs in a row: soccer ball + "BALL ACTION"; paw + "DUEL"; hamster + "MIGHTY MINI"; trophy + "VICTORY".
== MAIN GRID (8 panels, 2×4) ==
Each panel: thin white border, landscape 16:9. Number badge (1-8) top-left, teal "~1.5s" tag top-right, photorealistic still inside, bottom caption bar with icon + ALL-CAPS text.
1. [ball] "PET WORLD CUP FINAL KICKS OFF" — Wide stadium at golden hour. Midfield: orange tabby cat (red jersey) + golden retriever (blue jersey) facing off, ball between them. Deep background: enormous white goalposts with tiny hamster goalkeeper between them. Side: white rabbits in stands with pom-poms. Sunset cinematic light.
2. [paw] "RABBIT CHEERLEADERS BRING THE NOISE" — Close on stands: 3 fluffy white rabbits in tiny jerseys, mouths open cheering, paws shaking red and blue pom-poms in unison. Stadium lights, blurred crowd silhouettes behind.
3. [paw] "CAT VS DOG MIDFIELD DUEL" — Close-up: cat (red jersey) and dog (blue jersey) facing off, ball between their paws. Intense eye contact. Dirt kicking up. Whiskers and ear fluff in motion.
4. [ball] "DOG STEALS, BREAKS AWAY" — Dog with ball at paws charging through midfield. Cat sprinting behind, paws stretched, ears back.
5. [ball] "DOG WINDS UP — SHOT INCOMING" — Dog in classic kick pose, back leg lifted high, eyes locked on distant goal. Deep background: enormous white goalposts with TINY hamster goalkeeper barely visible — dramatic scale contrast.
6. [hamster] "HAMSTER VS THE GIANT GOAL" — Low angle close-up: enormous white goalposts towering over tiny golden Syrian hamster wearing mini yellow goalie gloves. Looking up determined, paws raised ready. Ball flying in from frame edge.
7. [hamster] "EPIC SAVE — TINY BUT MIGHTY" — Hamster in full sprawl leap, paws extended, blocking ball with tiny body. Mid-air freeze frame. Confetti falling, blurred crowd silhouettes raising arms.
8. [trophy] "HAMSTER HERO — TEAMS UNITE" — Wide final shot at golden hour: cat (red) and dog (blue) standing side by side staring in awe at tiny hamster holding up small golden trophy. White rabbits in stands raise pom-poms wildly. Confetti, golden sparkles.
== BOTTOM INFOGRAPHIC (4 sections horizontal) ==
[Clock] VIDEO FLOW: "8 shots × ~1.5s = 12s. Stadium → rabbit fans → cat-vs-dog rivalry → dog shoots → hamster saves → ensemble celebration."
[Camera] CAMERA TIPS: "Wide establishing on stadium, close-up on rabbits, low-angle on midfield duel, side-tracking on breakaway, scale frame on hamster vs goalposts, hero shot on save, wide on celebration."
[Sun] LIGHT & STYLE: "Photorealistic cinematic. Golden-hour stadium light throughout. Warm amber grass, lens flares, real DSLR/iPhone aesthetic. Crisp action focus."
[Person] CHARACTER NOTES: "Cast: orange tabby cat with M-mark (red jersey), golden retriever (blue jersey), golden Syrian hamster with mini yellow goalie gloves (yellow kit), 3-4 white fluffy rabbits with pom-poms. Hamster MUST appear dramatically tiny vs goalposts. All humans/crowd are faceless silhouettes — no faces. Each animal identical across panels."
== STRICT EXCLUSIONS ==
No real soccer team logos, no FIFA/Champions League/UEFA/World Cup branding, no real player faces, no sponsor logos, no national flags, no jersey numbers, no stadium signage with real text, no captions, no watermarks. Plain colored jerseys only. All 8 panels photorealistic — never cartoon/3D/illustration. All animals identical across panels.
27
Nine shot types from one image
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1536x1536 · job J15 · input: J13Result: Nine shot types with both figures consistent. Not asked for: a shot label and a tagline under each panel.
Works because. It names the 9 standard shot types and defines each (extreme long, long, American, medium, medium close-up, close-up, extreme close-up, worm's eye, bird's eye), with rules for groups and objects.
Carries the effect. “The same people/objects, same clothes, and same lighting across all 9 panels”, with the input image as the anchor.
To reuse. Keep the instruction block and replace the scene context at the end. The second half restates the grid in short form.
Sent for the image: The scene-context paragraph at the end (a mansion at dusk) was replaced with a description of the J13 campfire image, which is the input here.
Prompt, verbatim · 2,613 characters · 407 words
Show the full prose prompt
Given the scene context, generate the Cinematic Contact Sheet based on the instructions:
<instruction>
Analyze the entire composition of the input image. Identify ALL key subjects present (whether it's a single person, a group/couple, a vehicle, or a specific object) and their spatial relationship/interaction.
Generate a cohesive 3x3 grid "Cinematic Contact Sheet" featuring 9 distinct camera shots of exactly these subjects in the same environment.
You must adapt the standard cinematic shot types to fit the content (e.g., if a group, keep the group together; if an object, frame the whole object):
Row 1 (Establishing Context):
Extreme Long Shot (ELS): The subject(s) are seen small within the vast environment.
Long Shot (LS): The complete subject(s) or group is visible from top to bottom (head to toe / wheels to roof).
Medium Long Shot (American/3-4): Framed from knees up (for people) or a 3/4 view (for objects).
Row 2 (The Core Coverage):
4. Medium Shot (MS): Framed from the waist up (or the central core of the object). Focus on interaction/action.
5. Medium Close-Up (MCU): Framed from chest up. Intimate framing of the main subject(s).
6. Close-Up (CU): Tight framing on the face(s) or the "front" of the object.
Row 3 (Details & Angles):
7. Extreme Close-Up (ECU): Macro detail focusing intensely on a key feature (eyes, hands, logo, texture).
8. Low Angle Shot (Worm's Eye): Looking up at the subject(s) from the ground (imposing/heroic).
9. High Angle Shot (Bird's Eye): Looking down on the subject(s) from above.
Ensure strict consistency: The same people/objects, same clothes, and same lighting across all 9 panels. The depth of field should shift realistically (bokeh in close-ups).
</instruction>
A professional 3x3 cinematic storyboard grid containing 9 panels.
The grid showcases the specific subjects/scene from the input image in a comprehensive range of focal lengths.
Top Row: Wide environmental shot, Full body view, 3/4 cut.
Middle Row: Waist-up view, Chest-up view, Face/Front close-up.
Bottom Row: Macro detail, Low Angle, High Angle.
All frames feature photorealistic textures, consistent cinematic color grading, and correct framing for the specific number of subjects or objects analyzed.
Here is the scene context:
A wide cinematic exterior of the Whitmore estate at dusk in Buckhead, Atlanta. The massive gated modern mansion glows with warm amber lights against the dark tree line, with the circular driveway, lion fountain, and polished landscaping fully visible. The estate feels quiet, powerful, and untouchable — like a private world hidden behind wealth and security.
gpt-image-2.5-sunburst · high · 1024x1536 · job J47Result: Both sections, 8 panels with red borders, blue arrows and camera notes. The boy's sheet has no back view. Not asked for: a garbled tagline (“A small step can br bring. big wen friends.”) and slogans (“Same field. A kinder tomorrow”, “GOOD PEOPLE BETTER DAYS”).
Works because. Section A (5 characters: views, expressions, poses) and Section B (8 story beats) share one STYLE block. The sheet at the top acts as a reference for the panels below in the same image.
Carries the effect. “create fully original characters. Do not copy any existing anime...”. Useful when the style (“late 90s anime movie”) pulls towards known characters.
Watch out. 5 characters in 4 views plus 8 panels is a lot for one image; small faces will vary. For better consistency, make the sheet first, then the storyboard with the sheet as an input image.
Prompt, verbatim · 3,335 characters · 512 words
Show the full prose prompt
"Create a single vertical anime development board for an original emotional short film titled "One More Player". Output ONE combined image with 2 sections: top section = character design sheet, bottom section = cinematic storyboard page.
STYLE: ultra high-quality nostalgic late 90s / early 2000s anime movie look, premium hand-drawn animation style, beautifully painted backgrounds, rich environmental detail, cinematic composition, soft but highly refined shading, natural character anatomy, expressive faces, subtle texture, realistic wet ground reflections, warm cloudy sunset lighting, atmospheric depth, polished old-anime film aesthetic, visually stunning frame quality, clean studio-level pre-production presentation. Make it feel like a premium old anime film, not modern glossy anime.
IMPORTANT: create fully original characters. Do not copy any existing anime, manga, game, or sports anime characters, faces, hairstyles, outfits, or famous designs.
SECTION A: CHARACTER DESIGN SHEET
Show 5 original characters clearly and consistently.
Main child:
A shy little boy around 8 years old with short messy dark hair, large warm eyes, slim build, oversized yellow T-shirt, blue shorts, and small sneakers. He feels nervous at first, then happy and accepted.
Show:
- front view
- side view
- back view
- 3/4 view
- expressions: nervous, scared, surprised, soft smile, happy
- poses: holding soccer ball, standing shyly, running with ball
Teen group:
Four teenage boys around 14 to 16, sporty and friendly, each with different hairstyles and casual soccer outfits. They should look slightly intimidating at first from the little boy’s perspective, but actually kind and welcoming. One teen is the warm leader.
Show for the teen group:
- simple front, side, back, and 3/4 views
- relaxed poses
- soccer poses
- smiling expressions
Keep them original, visually distinct, and cohesive.
Add small handwritten design notes and simple color swatches.
SECTION B: STORYBOARD PAGE
Create 8 cinematic storyboard panels in a clean grid. Use red panel borders, blue motion arrows, handwritten camera notes, timing notes, and short action notes. Keep all characters consistent.
Story beats:
1. Four teenagers playing soccer on a neighborhood ground after light rain, warm cloudy sunset, wet reflections on the field.
2. The soccer ball flies away toward the sitting area beside the field.
3. A shy little boy sitting alone notices the ball and picks it up.
4. The teenage boys walk toward him together. From the little boy’s view, they look a little scary.
5. Close-up of the little boy holding the ball with a nervous face.
6. The warm leader smiles kindly and gestures with his hand, inviting the boy to come play.
7. The little boy’s face changes from scared to relieved, then he smiles and runs toward them.
8. Final wide shot: the little boy happily plays soccer with the four teens under warm sunset light.
ENVIRONMENT:
Neighborhood soccer ground, wet dusty field, simple goalpost, sitting area beside the field, chain-link fence, trees, cloudy sunset sky, calm old-anime mood.
FINAL GOAL:
Make this a clean, highly polished, premium-quality anime pre-production board with beautiful cinematic presentation, emotional storytelling, strong readability, and a heartwarming ending. The image should look visually rich, refined, and high-end."
gpt-image-2.5-flare · medium · 1152x1536 · job J48Result: The model chose a 2x3 grid in frame order. The same suit and sneakers in every frame.
Works because. Each frame starts with a shot type and ends with the same wardrobe line: “full beige suit, white t-shirt, white sleek sneakers”. The repetition is the consistency mechanism.
Watch out. It never says whether to make one grid or six images. For six calls, send the first result as a reference image with frames 2-6.
Style by name. “Peter Lindbergh influence” names a real photographer. Describing the look (black-and-white or muted tones, natural skin, minimal retouching) is more reliable and avoids the name. “f/1.2” on an extreme wide shot is a cue, not a simulation.
Sent for the image: Verbatim, as one call. The prompt does not say grid or six images, so the model chose.
Prompt, verbatim · 3,046 characters · 416 words
Show the full prose prompt
Frame 1 (Top-Down Establishing)
Extreme top-down shot in futuristic white curved hall, subject centered but compressed against smooth reflective floor, body aligned straight but slightly angled, arms close to body with subtle tension, head tilted upward toward camera, gaze directed straight up, strong circular architectural lines framing composition, soft studio lighting creating gentle gradients, shallow depth of field at f/1.2, smartphone-like proximity despite height, clean minimal atmosphere, dynamic opening frame, subtle athletic readiness, wearing full beige suit with white t-shirt and white sleek sneakers, Peter Lindbergh influence
Frame 2 (Low Angle Power)
Extreme low angle from floor level, subject towering above camera, legs forming strong base, torso slightly leaning forward, shoulders squared, arms slightly away from body, head angled downward, gaze into lens, strong vertical distortion, reflective floor amplifying silhouette, soft studio lighting with controlled highlights, shallow depth f/1.2, powerful dominance frame, subtle athletic tension, full beige suit, white t-shirt, white sleek sneakers, Peter Lindbergh influence
Frame 3 (Wide Isolation)
Wide long shot, subject very small within vast futuristic hall, surrounded by curved white architecture, large negative space dominating composition, subject standing still, body straight, arms relaxed but structured, head slightly turned, gaze outward into space, minimal reflections, soft even lighting, shallow depth f/1.2 isolating subject despite distance, cinematic silence, full beige suit, white t-shirt, white sleek sneakers, Peter Lindbergh influence
Frame 4 (Dynamic Close-Up Tilt)
Close-up with strong diagonal tilt, subject leaning into frame, shoulders sharply angled, one arm crossing body, head tilted sideways, gaze slightly off camera, tight crop cutting part of body, soft lighting wrapping contours, shallow depth f/1.2 isolating face plane, intense proximity and tension, full beige suit, white t-shirt visible, Peter Lindbergh influence
Frame 5 (Motion Step Forward)
Mid shot, subject stepping forward dynamically, one leg extended, weight shifting, arms slightly swinging with controlled motion, shoulders rotated, head facing forward, gaze ahead, slight motion blur on edges, futuristic curved walls passing behind, soft studio lighting, shallow depth f/1.2, transitional energy frame, full beige suit, white t-shirt, white sleek sneakers, Peter Lindbergh influence
Frame 6 (Final Grounded Frame)
Full body frontal shot, subject standing firmly, legs slightly apart, arms held with controlled tension, shoulders squared, head facing forward, gaze into camera, symmetrical composition slightly offset, minimal futuristic background, soft lighting emphasizing structure, shallow depth f/1.2, strong final impact frame, subtle athletic readiness, full beige suit, white t-shirt, white sleek sneakers, Peter Lindbergh influence
Negative Commands (apply to all frames):
no text, no watermark, no low quality, no cartoonish style
Aspect Ratio: 3:4
gpt-image-2.5-sunburst · high · 1536x864 · job J49Result: 6 panels in order. Every string is correct except panel 5, which ends 胸部コアが始まる instead of 胸部コアが発光する. The 最終認する typo was copied as written. Faces are hidden by bangs and a downward look. Not asked for: “SOLJEANNE PROTECTOR UNIT FOR A BRIGHTER TOMORROW”, “SCIENCE COURAGE A BRIGHTER TOMORROW”, “HEROES PROTECT A BRIGHTER TOMORROW”.
Works because. Header, 6 sections (title, position, caption, image description), footer and a grid spec. The order is explicit: inner suit, chest, waist, helmet prep, helmet on, done.
Carries the effect. The armor is specified once in subject and applies to every panel; the face is anonymized in every panel without the helmet.
Typo in the source. Step 1's caption reads 最終認する; it should be 最終確認する. The model renders text as written, so fix typos before sending.
To reuse. Any how-to board (cooking, assembly): keep layout.grid and text_rendering, rewrite the sections.
Sent for the image: Verbatim, including the 最終認する typo, to see if the model copies it.
Prompt, verbatim · 3,556 characters · 312 words
Show the full JSON prompt
Line breaks added for reading. Copy gives the original single line.
{
"type": "Japanese sci-fi armor dressing-process infographic",
"style": "cinematic live-action tokusatsu-inspired promotional board, realistic industrial lighting, polished metal surfaces, sharp photographic detail",
"theme": "manual pre-battle suit-up sequence for a female hero in a red, silver, black, and blue protector suit",
"subject": {
"character": {
"gender": "female",
"age": "young adult",
"identity": "helmetless heroine during assembly, face intentionally obscured or anonymized in every unhelmeted panel",
"hair": "dark brown to black hair tied in a high ponytail with bangs",
"undersuit": "glossy black skintight inner suit with silver chest panel and white neck ring",
"armor": "retro-futuristic protector armor with red shoulder and arm plates, silver breastplate and torso plating, circular blue chest core, red waist unit, white gloves, red forearm guards with yellow stripe accents",
"helmet": "round red-and-silver helmet with black visor"
},
"environment": {
"location": "high-tech industrial hangar or armor bay",
"background elements": [
"metal framework",
"robotic equipment",
"tool benches",
"armor racks",
"computer monitors",
"workshop lighting",
"bay corridor marked BAY-07 in final panel"
]
}
},
"layout": {
"header": {
"count": 2,
"labels": [
"ソルジャンヌ・スーツ 手動装着プロセス",
"専用プロテクタースーツ『ソルジャンヌ』を、戦闘前に手動で装着する様子。各ユニットを確実に装着し、システムを起動する。"
],
"design": "wide black-to-red gradient banner across top, large bold white Japanese text, diagonal red accent"
},
"sections": [
{
"title": "1 インナースーツの確認",
"position": "top-left",
"count": 1,
"labels": [
"各部のセンサーとコネクタをチェック。戦闘に備え、身体の状態を最終認する。"
],
"image": "three-quarter view of the heroine in only the black glossy inner suit, looking down while checking or tightening a wrist connector"
},
{
"title": "2 胸部・肩部アーマーの装着",
"position": "top-center",
"count": 1,
"labels": [
"胸部ユニットと肩部プロテクターを装着。コネクタを接続し、ロックを固定する。"
],
"image": "mid shot with chest armor and red shoulder plates installed, heroine fastening the front torso area with both hands"
},
{
"title": "3 腰部ユニット・ベルトの固定",
"position": "top-right",
"count": 1,
"labels": [
"ウエストユニットを装着し、各部のロックを確認。可動部の動作チェックを行う。"
],
"image": "mid shot with torso armor completed, heroine tightening or checking the waist belt and side locks"
},
{
"title": "4 ヘルメットの準備",
"position": "bottom-left",
"count": 1,
"labels": [
"ヘルメットのバイザーと内部システムをチェック。ヘッドセットとの同期を確認する。"
],
"image": "heroine holding the red helmet in both hands at chest height, showing the glossy black visor"
},
{
"title": "5 ヘルメットの装着・システム起動",
"position": "bottom-center",
"count": 1,
"labels": [
"ヘルメットを装着し、直上のコネクタをロック。全身のシステムが起動し、胸部コアが発光する。"
],
"image": "heroine placing the helmet onto her head with both hands; blue chest core glowing brightly"
},
{
"title": "6 装着完了",
"position": "bottom-right",
"count": 1,
"labels": [
"全システムの最終チェックを行い、戦闘モードへ。ソルジャンヌ、出撃準備完了!"
],
"image": "full-body frontal hero pose in a futuristic corridor, fully suited with helmet on, arms relaxed at sides"
}
],
"footer": {
"count": 1,
"labels": [
"一つ一つの装着が、命を守り、力を引き出す。 ソルジャンヌの戦いは、ここから始まる。"
],
"design": "dark red cinematic footer strip with centered white Japanese slogan"
},
"grid": {
"rows": 2,
"columns": 3,
"panel_count": 6,
"panel_borders": "thin white dividers",
"number_badges": 6
}
},
"text_rendering": {
"language": "Japanese",
"font": "bold sans-serif headline with smaller sans-serif body text",
"colors": "white text on black, red, and white info bars; red numbered squares with white numerals"
},
"composition": "16:9 wide infographic board, six equal photo panels arranged in a 3-by-2 grid, each panel captioned below with a red numbered box from 1 to 6",
"lighting": "moody workshop lighting with metallic reflections and red accent lights, realistic shadows, cinematic sci-fi atmosphere"
}
Focused edits and stylization
Ten prompts that start from an image you upload: restyles, identity-locked edits and format conversions. The edits of a person start from the same fictional person (J01), each from the original, never chained. The one exception is the anatomy plate, which needs a full-body photo, made from J01 in one step (J25). The product and crowd edits use a packshot or a photo I generated for the purpose (J26, J27, J28). Each edit shows its input as a small thumbnail under the image.
YouArt's own example image for this category, on GitHub (not copied here): edit.webp
31
Identity reference in JSON
JSONEdit: needs an input imageTemplateSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1152x1536 · job J16 · input: J01Result: Identity kept; headphones, hand pose and red object as asked. She smiles instead of the intense gaze the prompt asks for.
Works because.subject.instruction makes the uploaded photo “the single source of truth” for identity; wardrobe, headphones, backdrop and light are all new. This is a restyle, not a local edit.
Carries the effect. The accessory is described by colour and parts (glossy lime-green earcups, black padded headband), and a “large bright red circular object” in the foreground gives the colour contrast.
Not valid JSON until filled. The slot tokens put unescaped quotes inside JSON strings, so the published text does not parse. Fill the slots with plain values and it does.
Watch out. The negative list names the failure (“identity change”, “different person”). Listing what to preserve (face shape, freckles, skin tone, apparent age) is more useful; see the preservation block in Editing.
Sent for the image: Template slots filled: subject reference is “input image 1”; both pose slots use the hand pose already given in pose_details.
Inputgpt-image-2.5-sunburst · high · 1024x1536 · job J17 · input: J01Result: Identity and the new expression are there. The hair is only a little longer than in the input, not the long hair the prompt asks for.
Works because. Sections of short bullet lists: Identity, Expression, Styling, Lighting, Composition. The Identity list names the parts of the face to keep.
Conflict to look for. Styling asks for “long dark hair”; the input person has shoulder-length hair. The result shows which instruction wins. A good edit prompt does not contradict its input.
Carries the effect. The composition bullets place the eyes (“around the upper-middle area of the image”), which fixes the common crop with the face too low.
Sent for the image: Verbatim. Note the conflict: the prompt asks for long dark hair, the input has shoulder-length hair.
Use the uploaded portrait as the identity reference for the subject's face, hairstyle, facial structure, skin tone, expression, and overall impression.
Create a high-quality realistic emotional portrait of the same person with a soft sad mood and visible tears.
Core concept:
- a delicate, emotionally touching close-up portrait
- the subject looks quietly sad, as if holding back emotions
- the mood should feel fragile, intimate, soft, and beautiful
- the image should feel like a polished Korean-style emotional portrait
Identity:
- preserve the subject's recognizable identity
- keep the same face shape, eyes, nose, lips, jawline, hairstyle, and overall vibe
- do not make the face look generic or overly different
- keep the beauty natural and believable
Expression:
- slightly sad expression
- soft watery eyes
- one or two visible tear streaks running down the cheek
- lips softly closed or slightly parted
- emotional but restrained, not exaggerated
- the sadness should feel quiet, longing, and delicate
Styling:
- long dark hair with soft natural texture
- a few loose strands falling across the face
- clean natural makeup
- luminous skin
- simple dark top or minimal clothing visible
- overall styling should remain clean and understated so the face is the focus
Lighting:
- soft, moody lighting
- gentle highlights on the eyes, nose, lips, and tear tracks
- subtle shadow depth
- dark or muted background
- the lighting should feel intimate and cinematic, not harsh
Composition:
- vertical portrait composition
- close-up framing
- place the face in the upper half of the frame
- the center of the face should sit slightly above the vertical midpoint
- the eyes should fall around the upper-middle area of the image
- avoid placing the face too low in the frame
- keep the composition visually balanced and elegant
Mood and style:
- Korean emotional beauty portrait
- soft, melancholic, dreamy, intimate
- elegant and photogenic
- emotionally expressive without looking dramatic
- beautiful but slightly heartbreaking
Important visual priority:
- preserve identity clearly
- the face must remain the main focus
- tears should be visible but subtle
- expression should feel naturally sad and emotionally convincing
- the portrait should look beautiful, soft, and emotionally immersive
Negative prompt:
- no exaggerated crying
- no distorted face
- no cartoon style
- no anime style
- no harsh flash
- no messy background
- no over-retouched plastic skin
- no exaggerated smile
- no low-resolution image
- no text
- no watermark
33
Subject rebuilt from stone shards
JSONEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1152x1536 · job J18 · input: J01Result: Body, hair and mug are stone shards. The face stays close to photographic skin with seams, not slate.
Works because. The prompt string is a complete prose prompt. Material (slate and sandstone), gaps (“small visible gaps between shards (no pieces merging)”), light and background are physical facts.
Does nothing as a number.params (style_strength: 0.75, identity_preservation: 0.9) are not API parameters. The model reads them as words; they may nudge, but not as numbers.
Watch out. Stone is heavy stylization and eats identity. Judge the likeness by eye; face-matching tools are unreliable on stylized faces.
Sent for the image: The {{USER_IMAGE}} placeholder replaced with “input image 1”.
Prompt, verbatim · 1,986 characters · 228 words
Show the full JSON prompt
{
"task": "image_to_image_style_transfer",
"input_image": "{{USER_IMAGE}}",
"prompt": "Create a hyper-real 3D studio composition that recreates the main subject from the provided image as a fragmented stone assemblage. The subject must be built from separate, clearly detached rock pieces with small visible gaps between shards (no pieces merging). Material look: fragmented slate + sandstone shards with chiseled edges, crisp fractures, visible stone grain, micro-scratches, and realistic roughness. Color palette: predominantly dark slate with subtle warm-ochre sandstone accents. Lighting: soft studio key light from top-left, gentle fill, subtle contact shadows under each shard, realistic ambient occlusion in crevices, clean reflections kept minimal. Background: minimal off-white seamless backdrop, no texture. Framing: centered, clean, straight-on, subject fully readable. Add a few tiny debris chips floating or resting near the base for depth. Preserve the subject’s identity, proportions, and recognizable silhouette from the input image while transforming all surfaces into stone fragments. Hyper-real, high detail, sharp focus, 8k render quality.",
"negative_prompt": "text, typography, logo, watermark, signature, extra props, busy background, fog, heavy bloom, cartoon, illustration, lowpoly, plastic, metal, glossy paint, melted shapes, merged fragments, unreadable subject, blur, noise, low resolution, oversharpening halos, distorted face/body, extra limbs, deformed geometry",
"output": {
"aspect_ratio": "use_input_aspect_ratio",
"background": "off_white",
"camera": {
"angle": "straight_on",
"framing": "centered",
"distance": "medium"
}
},
"params": {
"style_strength": 0.75,
"identity_preservation": 0.9,
"detail_level": "very_high",
"lighting_preset": "soft_studio_top_left",
"shadow_intensity": "subtle",
"gap_visibility": "clear",
"debris_chips": "few_tiny",
"no_text": true
}
}
34
Photo to anatomical atlas plate
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1024x1536 · job J50 · input: J25Result: An anatomy plate in the source pose, with the sweater and jeans as outlines. The Latin labels I checked are spelled correctly, with skull, spine, pelvis, hand and foot insets and the caption “Tabula Anatomica — Figura I / Studium Corporis Humani”. The face stays a photo instead of becoming a drawing. The scapula has no label.
Works because. A preserve list for the pose (limb position, torso rotation, head angle, gaze, hand placement) and a style described by materials: graphite lines, red and blue ink, watercolor washes, typewriter labels.
Carries the effect. The list of Latin terms gives the model real words to place, and “Avoid horror, gore, blood” keeps it a textbook plate.
Watch out. “slightly imperfect, with subtle ink bleed, uneven typewriter spacing” invites misspelled labels. Check the Latin.
Sent for the image: Verbatim, with a full-body photo of the test person (J25, made from J01) as the input.
Edited to pass 2.5 moderation. As written, this prompt was refused twice by GPT Image 2.5 Sunburst output moderation (category: sexual) in my run on 2026-09-26. The image comes from the third of three minimal rewrites, the first one that passed. Changes (removed › added):
Clothing and accessories may be simplified into faint translucent contour lines so the anatomical structures remain readable. › Keep her sweater and jeans as faint translucent contour lines over the anatomy so the anatomical structures remain readable.
Musculus pectoralis major, › (deleted)
Musculus gluteus maximus, › (deleted)
reveals bones, superficial muscles, tendons, ligaments, and selected organ silhouettes › reveals bones, superficial muscles, tendons and ligaments
Rewrite 1 (only the clothing line: keep her sweater and jeans as contour lines) and rewrite 2 (also dropping the pectoralis and gluteus labels) were refused. Rewrite 3 also dropped the organ silhouettes and passed.
The prompt below is the verbatim original.
Prompt, verbatim · 3,279 characters · 441 words
Show the full prose prompt
Edit the provided image into an antique anatomical book plate. Preserve the original character's pose, camera angle, body proportions, silhouette, gesture, and overall composition as closely as possible, but transform the subject into a refined anatomical study illustration. The final image should look like a vintage medical anatomy textbook page printed on aged ivory paper, with sepia stains, subtle paper fibers, worn edges, faint foxing marks, and archival ink texture.
Convert the visible body into a clean anatomical sketch that reveals bones, superficial muscles, tendons, ligaments, and selected organ silhouettes in a didactic, non-gory, museum-quality style. Keep the anatomy elegant and educational, like a 19th-century anatomical atlas mixed with precise modern medical illustration. Use thin graphite lines, red and blue anatomical ink accents, muted watercolor washes, and delicate hatching. Avoid horror, gore, blood, injury, mutilation, or surgical violence.
The figure must remain in the same pose as the source image, including limb position, torso rotation, head angle, gaze direction, hand placement, and perspective distortion. If the original image is anime, game art, illustration, or realistic photography, reinterpret it consistently as an anatomical diagram while keeping the recognizable posture and visual identity cues only as subtle external outlines. Clothing and accessories may be simplified into faint translucent contour lines so the anatomical structures remain readable.
Add anatomical labels in Latin using a vintage typewriter font, as if typed onto the page. Labels should be connected to structures with thin black leader lines. Use accurate Latin anatomical terminology, including examples such as: Cranium, Mandibula, Clavicula, Sternum, Costae, Scapula, Humerus, Radius, Ulna, Vertebrae cervicales, Vertebrae thoracicae, Vertebrae lumbales, Pelvis, Os ilium, Os ischii, Femur, Patella, Tibia, Fibula, Talus, Calcaneus, Musculus sternocleidomastoideus, Musculus trapezius, Musculus pectoralis major, Musculus deltoideus, Musculus biceps brachii, Musculus rectus abdominis, Musculus obliquus externus abdominis, Musculus iliopsoas, Musculus gluteus maximus, Musculus sartorius, Musculus quadriceps femoris, Musculus biceps femoris, Musculus gastrocnemius, Ligamentum inguinale, Articulatio coxae, Articulatio genus, Articulatio talocruralis.
Place the labels naturally around the figure, balancing readability and visual elegance. The text should look authentic, slightly imperfect, with subtle ink bleed, uneven typewriter spacing, and small alignment irregularities. Include a small archival caption at the bottom of the page, such as “Tabula Anatomica — Figura I”, “Studium Corporis Humani”, or “Atlas Anatomicus Vetus”. Do not use Japanese labels unless specifically requested; prioritize Latin anatomical nomenclature.
Lighting should be soft and flat like a scanned antique page. The final result must feel like a scholarly anatomical illustration, not a modern infographic. High detail, precise linework, aged paper texture, anatomical accuracy, elegant composition, subtle red muscle fibers, pale blue veins, ivory bone rendering, dark graphite outlines, old medical atlas atmosphere, realistic printed-page imperfections.
35
Photo on top, storybook drawing below
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1152x1536 · job J19 · input: J01Result: Photo and drawing match. The caption is the sample text copied as is, “little happy girl ♡”, although the prompt says to adapt or drop it when the subject is not a girl.
Works because. A REFERENCE PRIORITY list ranks what matters, subject and pose first, decorations last. Ranking tells the model what to give up first when it cannot keep everything.
Carries the effect. “Never sacrifice subject accuracy for artistic stylization”, plus a list of what must match between the photo and the drawing.
Watch out. The sample caption is “little happy girl ♡”, with a rule to adapt it. The input here is an adult woman, so the result shows whether the rule is followed.
Also in the repo. A near-duplicate of this prompt by @MahnoorAi12.
Prompt, verbatim · 3,696 characters · 531 words
Show the full prose prompt
Create a single vertical 3:4 PHOTO + HAND-DRAWN ILLUSTRATION artwork based ONLY on the uploaded photograph.
Use the uploaded photograph as the ONLY visual reference. Analyze the photograph independently and reproduce the exact subject, pose, expression, clothing, hairstyle, accessories, objects, proportions, and overall moment in the illustration. Do not introduce elements from any previous image or reference.
TOP — ORIGINAL PHOTOGRAPH
Use approximately the upper 50% of the canvas for the original photograph.
Preserve the photograph faithfully and naturally:
- Keep the original subject exactly recognizable.
- Preserve the face, expression, hairstyle, clothing, accessories, pose, body proportions, and visible objects.
- Preserve the original environment and composition.
- Do not retouch, beautify, replace, or redesign the subject.
- Keep the photograph realistic with its original lighting, colors, textures, and details.
BOTTOM — HAND-DRAWN ILLUSTRATION
Use approximately the lower 50% for a charming handmade illustration of the SAME photograph.
Transform the photographic moment into a delicate children's storybook-style hand-drawn artwork.
Illustration style:
- Soft watercolor and colored-pencil texture
- Gentle pencil/ink outlines
- Slightly imperfect handmade strokes
- Warm, tactile off-white textured paper
- Soft natural shading
- Subtle watercolor washes
- Cute, expressive, wholesome appearance
- Delicate handcrafted imperfections
- No glossy digital-rendered appearance
- No photorealism
- No 3D CGI
- No anime or cartoon franchise style
Most importantly, accurately preserve the subject from the photograph:
- Same pose and body position
- Same facial expression
- Same hairstyle
- Same clothing and clothing patterns
- Same shoes/accessories
- Same proportions
- Same distinctive visual details
- Same overall action or mood
Simplify the background substantially. Do not recreate every background detail. Keep only a very subtle suggestion of the original environment, allowing the illustrated subject to remain the main focus.
Place the illustrated subject slightly smaller than the photograph's subject, surrounded by generous negative space.
MINIMAL DECORATIVE DETAILS
Add a few tiny hand-drawn decorative marks around the illustration, such as:
- small hearts
- simple motion lines
- tiny doodles
- subtle stars or flowers
Keep these decorations sparse and tasteful.
Add a small handwritten caption near the illustration, such as:
"little happy girl ♡"
The caption should look naturally handwritten with a casual brown/gray pencil or ink texture.
If the uploaded subject is not a girl, automatically adapt the wording appropriately or omit the caption entirely.
OVERALL AESTHETIC
The final artwork should feel like a precious memory transformed into a page from a handmade children's picture book.
Composition:
- Vertical 3:4
- Clean separation between photograph and illustration
- Soft warm editorial presentation
- Large amount of negative space around the illustration
- Minimal, elegant, sentimental
- Premium handmade stationery / illustrated scrapbook aesthetic
CRITICAL RULE:
The uploaded photograph is the sole source of truth. Do not invent, substitute, merge, or carry over any person, clothing, object, background element, color, pose, or detail from another image.
The bottom illustration must clearly look like a hand-drawn version of the EXACT moment shown in the top photograph.
REFERENCE PRIORITY:
1. Exact subject and pose
2. Facial expression and hairstyle
3. Clothing and accessories
4. Overall composition
5. Illustration style
6. Minimal decorative elements
Never sacrifice subject accuracy for artistic stylization.
The JSON as written, text to image · gpt-image-2.5-flare · medium · 1536x1152 · job J51Result: Text to image. Headline, Source/Output labels and both captions exactly as written. The overhead and ground-level panels match the described figures (maroon figure, crouched green figure, woman in blue).InputThe real edit, on my overhead photo · gpt-image-2.5-flare · medium · 1536x1024 · job J52 · input: J28Result: The people are kept remarkably well: the vendor with scarf and paper bag, the man in a brown jacket and flat cap, the man in navy with a basket, the gray-haired woman, the man in a fedora, the horse and the cheese stall. The camera ends up at standing eye height, not ground level.
What it is. A prompt for a comparison graphic that draws both halves, the “source” and the “output”, from text. It does not change an existing image.
For a real POV change. Send your image with one instruction: “Move the camera POV to be at ground level in the crowd. Keep the people, clothing, carriage and light.” The first sentence is this prompt's headline text, and it is the useful part. The second image here is that edit (J52), with “carriage” changed to “cart” to match my photo.
Useful anyway. For demo graphics: text_elements plus two images with a role each.
Sent for the image (J51): Template slot filled with its default. Text to image: the JSON describes both panels, so no input was sent.
Sent for the image (J52): The real edit the notes suggest: my overhead market photo (J28) as input, with the headline sentence plus one keep line.
Prompt, verbatim · 2,546 characters · 269 words
Show the full JSON prompt
Line breaks added for reading. Copy gives the original single line.
{
"type": "comparison graphic",
"style": "anime cinematic demonstration image on a black presentation background",
"canvas": {
"aspect_ratio": "4:3",
"background": "solid black"
},
"text_elements": [
{
"text": "{argument name=\"headline text\" default=\"Move the camera POV to be at ground level in the crowd.\"}",
"position": "top center",
"style": "large white sans-serif"
},
{
"text": "Source",
"position": "left of upper image",
"style": "large white sans-serif"
},
{
"text": "Output",
"position": "left of lower image",
"style": "large white sans-serif"
}
],
"layout": {
"sections": [
{
"title": "Source",
"position": "upper center",
"count": 1,
"labels": [
"overhead crowd scene"
]
},
{
"title": "Output",
"position": "lower center",
"count": 1,
"labels": [
"ground-level crowd POV scene"
]
}
],
"image_frames": 2
},
"images": [
{
"role": "source image",
"composition": "busy top-down view of a densely packed historical street crowd, seen from above",
"scene": "a chaotic crowd gathered around a wagon and a horse-drawn carriage, people pressed shoulder to shoulder, many wearing caps and muted early-20th-century or old-European clothing, bundles and sacks visible, one brown horse at the right edge, wooden wagon wheel and cart structure partially visible",
"camera": "high overhead bird's-eye angle looking down into the crowd",
"lighting": "soft daylight",
"color_palette": "muted earthy browns, dusty blues, beige, olive, warm gray",
"rendering": "hand-painted anime film still, detailed crowd illustration, slightly soft shading"
},
{
"role": "output image",
"composition": "the same crowded historical street reimagined from inside the mass of people at near-ground height",
"scene": "view from within the crowd beside a carriage wheel, bodies filling the foreground and midground, a person in dark maroon clothing bent forward at left, a crouched figure in green near the bottom center, a woman in a light blue dress at right-center turning back, tightly packed figures, horse and cart implied nearby, dramatic sense of compression and closeness",
"camera": "very low ground-level POV from inside the crowd, upward and forward through people, emphasizing complex occlusion and depth",
"lighting": "soft daylight with warm cinematic shadows",
"color_palette": "muted earthy browns, dusty blues, beige, olive, warm gray",
"rendering": "hand-painted anime film still, cinematic perspective shift, detailed character crowding, soft painterly shading"
}
],
"overall_goal": "show a before-and-after camera angle transformation of the same anime crowd scene, with the output moving from an overhead view to a low immersive POV inside the crowd"
}
37
Any image to paint-by-numbers
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1248x1760 · job J20 · input: J06Result: Clean line art with numbered regions and a legend. The legend jumps from 8 to 13, so numbers 9-12 are missing, and the J06 labels stay as text.
Works because. Hard print constraints: white background, black outlines only, a number in every closed region, a legend at the bottom. The colour count adapts to the image, from 1 to 14.
Watch out. “A4 portrait ratio canvas at 300 DPI” is not a pixel setting. A4 at 300 DPI would be 2480x3508, above the 2560x1440 limit where sizes become experimental; this test used 1248x1760.
Check. The hard part is numbers matching the legend. Spot-check a few regions.
Sent for the image: Only the text after “Prompt below 👇” was sent. The row starts with the author's post text, which is not part of the prompt.
Prompt, verbatim · 2,568 characters · 403 words
Show the full prose prompt
🎨 This GPT Image 2 prompt turns any image into a printable paint-by-numbers page. For best results, use ChatGPT in Thinking mode. I think we should call it numberfying. The intellectuals can call it Colorbookification. 😂 I tried it with a few TurboToadToken images and the results are wild. 🐸💛 Create it. Print it. Paint it. For kids, adults, and anyone who wants an easy creative activity. Try it and post your results in the comments. Prompt below 👇 Create a professional high-resolution paint-by-numbers template from the provided image as a single PNG only. Use an A4 portrait ratio canvas at 300 DPI with a pure white background. The main artwork must be a clean black-and-white line drawing containing only black outlines, black numbers, and white unfilled paintable areas. Do not create a PDF, document, mockup, multi-page layout, or colored version. Do not use color, gray, shading, texture, gradients, transparency, or pre-filled areas in the main artwork. Analyze the original image and choose the number of colors adaptively based on image complexity, from 1 to a maximum of 14 colors. If the image has only 1-2 important colors, use only 1-2 colors; for simple images use 3-6 colors; for medium complexity use 7-10 colors; for complex, photographic, or highly detailed images use 11-14 colors. Do not force unnecessary colors. Assign each selected color one unique number and merge similar shades unless separation is essential for recognition. Convert the image into simplified, fully closed, easy-to-paint regions while preserving the recognizable composition, main objects, characters, text, logos, foreground, background, and key details. Every paintable white region must contain the correct black number, centered clearly inside the matching area, readable, upright, and not touching outlines. Large black or dark areas from the original must remain white paintable regions with the correct number, not filled black. Only tiny permanent black line details, such as mouth lines, nostrils, pupils, eyelashes, seams, or small decorative strokes, may remain unnumbered. At the bottom of the same PNG, add a neat legend section with the heading "Instructions", the sentence "Paint the areas with the matching colors.", and a list of every used number with a small color swatch and the English color name. Color swatches may appear only in the legend. Ensure the numbering exactly matches the legend, every paintable area is numbered, all regions are closed, and the final result is clean, professional, print-ready, and recognizable as the original image.
38
Catalog page: hero shot and technical drawings
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1152x1536 · job J53 · input: J26Result: Lamp shape and materials kept. Sepia drawings with invented dimensions (Ø 330 mm, 540 mm) and 3 captioned swatches. A real magazine title, Kinfolk, appears on a book spine; the one retry put it there again, so this is attempt 1.
Works because. Two sections with fixed roles: a lifestyle hero on top, a specification grid below with orthographic drawings and material swatches.
Carries the effect. The product lock in brackets (“preserving its form, proportions, materials, and identity without redesign”) and “no perspective distortion in the technical drawings”.
Watch out. The model will invent the measurement callouts. Replace them with real numbers afterwards.
Sent for the image: Verbatim, with my lamp packshot (J26) as the uploaded image.
Prompt, verbatim · 1,462 characters · 207 words
Show the full prose prompt
Create a vertical 3:4 product design catalog page with a warm neutral paper-like background.
Top section — lifestyle hero shot: place the product (use the uploaded image as the exact reference, preserving its form, proportions, materials, and identity without redesign) center-dominant with generous whitespace. Setting is a minimal architectural interior with a textured plaster wall and subtle concrete/stone floor. Lighting is natural sunlight angled from the side, soft but casting high-contrast shadows. Render in editorial lifestyle photography style, high realism, warm and muted color grading.
Bottom section — technical specification panel laid out in a clean modular grid:
- Bottom left and center: orthographic architectural line drawings showing front view, side view, and three-quarter cutaway/profile view. Lines in muted red or sepia, fine technical weight, with minimal editorial measurement and construction callouts.
- Bottom right: 3-4 material swatch samples derived from the product's actual materials (fabric, leather, metal, wood, or plastic as applicable), in square or rectangular format with small editorial captions.
Typography: minimal editorial style, subtle captions only, no large headlines, soft black or dark brown.
Overall mood: design catalog / product design journal — architectural, premium, calm. No clutter, no bold colors, no heavy branding, no decorative graphics, no perspective distortion in the technical drawings.
39
Pencil sketch holding a real mug
JSONEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1024x1792 · job J54 · input: J01Result: The J01 likeness holds, with the face in graphite. The jacket is pencil too, not full colour as asked; hand and mug are photoreal. KAVA and K.A.V.A as written. The steam rises in front of her chin instead of covering the lower face. Not asked for: “SAME PEOPLE BRIGHTER DAYS” on a notebook and “Good Coffee Better People” on a bag.
Works because. It says where each medium applies: face in graphite, clothing, hand and mug photoreal, steam covering part of the face. That is what makes the trompe-l'oeil read.
Does nothing."size", "quality": "ultra" and "detail_level": "8K" are text here. The API sets size and quality as parameters, and “ultra” is not one of its quality values.
To reuse. “handsome young man with ... glasses, smile” describes the author's reference person. With another input, rewrite these words to match it.
Sent for the image: The man became a woman to match the input (J01): “handsome young man ... glasses” became “young woman”, he/his/man became she/her/woman. The mug label “AEGON” (a real insurer) became “KAVA”, and the signature “A” / “A.E.G.O.N” became “K” / “K.A.V.A”. Size from the JSON (1024x1792).
Prompt, verbatim · 2,280 characters · 315 words
Show the full JSON prompt
{
"prompt": "A hyper-realistic top-down photograph of a white sketch paper lying on a rustic wooden table. On the paper is an incredibly detailed graphite pencil sketch of a handsome young man with the exact facial features, hairstyle, glasses, smile, proportions, and expression from the uploaded reference image. The portrait is rendered in realistic graphite pencil shading with exceptional detail. He is wearing a modern black American-style jacket and shirt, with only the clothing rendered in full color while the face remains a pencil sketch. The man is holding a real photorealistic 3D black coffee mug labeled 'AEGON' that appears to break through the boundaries of the paper and extend into the real world. The coffee mug is rendered in full color with realistic reflections and texture. His hand holding the mug is also fully photorealistic and colored, seamlessly blending with the sketch to create an impressive trompe-l'oeil illusion. Realistic steam rises from the hot coffee and partially covers the lower part of his face. The white sketch paper rests on a rustic wooden tabletop surrounded by scattered coffee beans, two sharpened black pencils, subtle sketching tools, and warm natural sunlight creating soft shadows. Cozy coffee shop atmosphere, shallow depth of field, cinematic lighting, premium composition. Photorealistic mixed-media artwork, pencil drawing combined with 3D realism, illusion art, trompe-l'oeil effect, ultra-detailed textures, realistic steam, professional studio-quality lighting, highly detailed graphite shading, 8K resolution, masterpiece quality. In the bottom-left corner of the paper, add a small minimalist triangle containing a stylized letter 'A' with the text 'A.E.G.O.N' elegantly written beneath it as an artist signature. Preserve the reference face exactly.",
"size": "1024x1792",
"style": "photorealistic",
"quality": "ultra",
"aspect_ratio": "9:16",
"lighting": "warm natural sunlight",
"camera_angle": "top-down",
"detail_level": "8K",
"negative_prompt": "blurry, low resolution, distorted face, extra fingers, duplicate hands, bad anatomy, cartoon, anime, painting, watercolor, oversaturated colors, low detail, text artifacts, cropped face, unrealistic steam, deformed mug, unrealistic proportions"
}
gpt-image-2.5-flare · medium · 1152x1536 · job J55Result: 7 motifs (lake, stag, oak leaves, cabin, capercaillie, porcini, castle ruin) on bright white paper, with no yellow cast. Not asked for: a title block, “Bohemian Forest / CZECH REPUBLIC / Forests · Mountains · Wildlife”, probably prompted by “Language: English”.
Works because. It describes a style by medium and rules, not by an artist's name: gouache, risograph and silkscreen texture, flat shapes, at most 8-10 elements, a named palette.
Carries the effect. “crisp neutral white balance ... no yellow tint no beige cast no sepia tone” counters the warm cast these models add to paper textures.
To reuse. Replace [Subject] in all three places. “8K ultra-high resolution” and “Language: English” are filler. This works as a house style across many subjects.
Sent for the image: [Subject] replaced in all three places with “the Bohemian Forest in autumn”.
Prompt, verbatim · 2,587 characters · 326 words
Show the full prose prompt
Create a premium mid-century-inspired flat editorial illustration of [Subject], rendered in a refined handcrafted printmaking aesthetic. The artwork should feel like a sophisticated vintage children’s encyclopedia illustration combined with Scandinavian editorial design and retro nature-book graphics.
Language: English 8K ultra-high resolution with a 3:4 aspect ratio
STYLE: minimalist flat-shape illustration soft geometric simplification handcrafted gouache + risograph + silkscreen print texture subtle paper grain and dry ink imperfections muted refined color palette clean matte colors crisp neutral white balance bright museum-quality paper appearance no yellow tint no beige cast no sepia tone no glossy rendering no photorealism no 3D shading elegant editorial composition premium picture-book atmosphere modern museum-shop illustration vibe
COMPOSITION: spacious balanced layout with multiple elements harmoniously arranged simplified organic shapes stylized plants, objects, animals, architecture, landmarks, or symbolic elements depending on [Subject] visual storytelling through clustered illustrated motifs maximum 8–10 illustrated elements total playful but refined asymmetrical editorial balance large clean negative space on soft neutral white paper background premium Scandinavian editorial layout feeling
TEXTURE & PRINT FEEL: subtle premium paper texture refined layered ink overlap delicate risograph misalignment soft dry brush grain light handcrafted print imperfections tactile handcrafted surface feeling clean bright paper surface controlled ink density airy and crisp visual atmosphere
COLOR SYSTEM: muted teal dusty blue sage green cool gray faded coral soft navy desaturated mineral tones restrained natural palette (adjust palette naturally to fit [Subject])
ILLUSTRATION RULES: simplified eyes and facial features rounded organic silhouettes minimal line usage flat layered color blocking decorative but restrained detail density avoid excessive realism avoid anime style avoid modern vector corporate style avoid bright neon colors avoid sharp digital gradients avoid yellowish paper tone avoid vintage sepia coloration
MOOD: quietly playful, educational, nostalgic, timeless, sophisticated, airy, clean, premium editorial illustration, Scandinavian museum-book elegance
OUTPUT: high-resolution editorial illustration, ultra-clean composition, refined handcrafted print aesthetic, bright neutral paper tones, crisp museum-quality editorial finish, suitable for premium magazine spreads, art books, museum posters, or luxury children’s publishing.
Character sheets
Nine sheets, weighted towards portraits, restyles of one person and identity edits. Counted views and a fixed layout are what make a sheet usable as a reference later.
YouArt's own example image for this category, on GitHub (not copied here): character-sheet.webp
gpt-image-2.5-sunburst · high · 1536x864 · job J21Result: Matches the prompt: 4 depictions, the counted details and outfit colours consistent across views, no text.
Works because. “Exactly 4 character depictions” with sizes and positions, then a “Visible counted elements” section: hair clip, ponytail tie, drawstrings, earrings, socks, sneakers.
Carries the effect. Small details are what break first between views. Counting them gives you a checklist to grade the result against.
To reuse. The slots are age and vibe, hair, outfit. Keep “preserve the same outfit colors and hairstyle in every view”.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 2,769 characters · 415 words
Show the full prose prompt
Goal: Create a polished 3D animated character design sheet for a cheerful young girl, suitable for animation reference.
Canvas: Wide horizontal 16:9 composition on a clean white studio background with soft even lighting, no visible floor line, no text, no logos, no watermark.
Layout: Show exactly 4 character depictions across the canvas: 1 large close-up portrait on the left occupying about 40% of the width, cropped from chest upward and facing slightly toward the viewer; 3 smaller full-body turnarounds on the right arranged left to right with even spacing: front view, three-quarter side view facing right, and back view.
Character details: The character is {argument name="character age and vibe" default="a cheerful preteen girl with a bright playful personality"}. She has {argument name="hair color and style" default="warm brown hair in a high messy ponytail with loose flyaway strands and side-swept bangs"}, large expressive brown eyes, soft rounded facial features, rosy cheeks, light freckles across the nose and cheeks, small purple hoop earrings, and an open happy smile in the close-up and front view. Add a colorful geometric triangular hair clip and a rainbow-striped ponytail tie.
Outfit: Dress her in {argument name="main outfit" default="an oversized color-block hoodie with a yellow hood and upper chest, cyan left sleeve and lower diagonal panel, purple right sleeve and diagonal panel, and a pink kangaroo pocket with pink drawstrings and dark cord tips"}. Pair it with navy athletic shorts with a vertical purple stripe, white crew socks with pastel colored stripes, and chunky colorful sneakers in pink, teal, yellow, purple, and white. Include a small purple wristband or watch on one wrist.
Visible counted elements: Exactly 4 character depictions total: large close-up portrait, full-body front view, full-body right-facing three-quarter view, full-body back view. Exactly 1 prominent triangular hair clip on each visible head where the front or side is visible. Exactly 1 ponytail tie. Exactly 2 hoodie drawstrings with dark tips. Exactly 2 earrings visible in the close-up if both ears are shown, otherwise at least 1 visible purple hoop earring. Exactly 2 socks and 2 sneakers on each full-body view.
Visual style: High-quality stylized 3D animation look, similar to a modern family-friendly animated film; soft subsurface skin shading, glossy lively eyes, detailed hair clumps, clean shapes, saturated candy colors, crisp character-sheet presentation, gentle shadows, professional model-turnaround lighting.
Constraints: Keep proportions consistent between all 4 depictions, preserve the same outfit colors and hairstyle in every view, use a white background only, do not add labels, props, scenery, extra characters, or text.
42
Tri-fold concept sheet with style and subject slots
gpt-image-2.5-sunburst · high · 1536x864 · job J22Result: Matches the layout: hero pose, 2 turnaround views, 3 expressions, 2 gear panels, 6 swatches. Name, age and theme are invented, as the prompt allows.
Works because. [STYLE] and [SUBJECT_DESCRIPTION] are defined at the top, then the layout is given in percentages (40/35/25) with counts: 2 turnaround views, 3 expressions, 2 gear panels, 5-6 swatches.
Carries the effect. “Apply [STYLE] only to the character and visual elements” keeps the layout clean while the character is claymation.
Watch out. “Infer all missing details” lets the model invent the name and the specs. Give them if you need them fixed.
Prompt, verbatim · 2,708 characters · 399 words
Show the full prose prompt
Create a high-end, asymmetric editorial CHARACTER CONCEPT SHOWCASE from these inputs:
[STYLE]: stylized 3D stop-motion claymation style with rich tactile textures of clay, felt, and leather, and warm cinematic studio lighting
[SUBJECT_DESCRIPTION]: A charming and slightly eccentric traveling medieval alchemist and cartographer. He wears a heavy, oversized patched wool coat over a worn leather tunic, multiple small glowing potion vials and rolled-up parchment scrolls strapped to his utility belt, a wide-brimmed traveler's hat, and thick round brass spectacles. He has a messy, hand-sculpted ginger beard, warm curious eyes, and a friendly smile. He carries an ancient, brass-trimmed leather satchel. His design features exaggerated, whimsical proportions and a cozy, rustic medieval aesthetic.
Create the layout in a clean 16:9 widescreen format on a neutral studio gray or warm off-white background with a minimal technical border. The design must look like a premium production visual bible, using clean typography, no clutter, no watermarks, and no logos. Apply [STYLE] only to the character and visual elements, keeping the presentation layout clean, structured, and minimal.
Infer all missing details from the subject description, including name, role, brief background specs, and a cohesive color palette.
Use this tri-fold layout:
1. HERO SPOTLIGHT (Left 40% of the board)
- Show one large, highly detailed full-body dynamic action pose of the subject.
- This pose should showcase the character's primary personality, attitude, and silhouette.
2. TECHNICAL TURNAROUND (Center 35% of the board)
- Show exactly two clean full-body views: Front View and Back View.
- The subject should be in a relaxed, neutral stance.
- Place these views over very subtle vertical and horizontal grid lines resembling a technical schematic blueprint.
3. KEY DETAILS (Right 25% of the board)
- EXPRESSION TRIO: Exactly 3 large, highly expressive close-up headshots showing core emotional states: Calm/Neutral, Highly Focused/Intense, and a Dynamic/Expressive emotion (like a smirk or fierce grin).
- GEAR CALLOUTS: Exactly 2 clean, isolated close-up panels showing primary wardrobe textures, signature accessories, or weapons/gear.
4. SPECS & COLOR BANNER (Bottom Edge)
- A minimalist, horizontal typography block listing Name, Role, Age, and Core Theme.
- Adjacent to the text, display 5 to 6 clean geometric color swatches showing the character's primary color palette with no labels.
Ensure complete character and costume consistency across all sections. The Hero Spotlight must visually anchor the sheet, offering a clean, open, and professional layout that avoids dense, repetitive, or cluttered grids.
gpt-image-2.5-sunburst · high · 1536x1024 · job J56Result: Name, age, classification and profile as written. Turnaround, 5 named expressions, wheel schematics, car specs and fictional patches (VULTEX, KARSH, RAVEN). The style is clean digital painting; the grease-pencil lines and impasto are faint.
Works because. [identity_module], [turnaround_module], [gear_module], [expression_module], [style_tags]: section tags inside one paragraph, like JSON keys without braces.
Carries the effect. Very specific wardrobe (oxblood bomber, locs with golden cuff beads, fraying ivory hand-wrap tape) and five named expressions, each with its facial mechanics.
Style by name. “Arcane style” names a TV series. The prompt already describes the look (2.5D, grease-pencil outlines, impasto); the name can go.
Prompt, verbatim · 3,652 characters · 510 words
Show the full prose prompt
[layout_setup]: A comprehensive, full-page character model design sheet layout strictly on a pristine solid white background with no gradients, no environmental art, and no photorealistic elements whatsoever. [identity_module]: On the left side, large ultra-bold minimalist design typography spelling 'REINA VOSS' next to clean monospace text columns detailing character age 24, classification traits as underground circuit legend and drift specialist, tactical description blocks reading former getaway wheelman turned undefeated canyon queen with a photographic memory for road geometry and an obsessive ritual of wrapping her knuckles before every race. [turnaround_module]: Centered prominently on the sheet is a full-body orthographic turnaround lineup showing identical front view, side profile view, and back view of a tall, sharp-jawed young woman with deep terracotta skin and dense voluminous shoulder-length locs gathered loosely at the nape with a worn leather cord, a few golden cuff beads threaded throughout, wearing a cropped vintage racing-inspired bomber jacket in faded oxblood leather with oversized wool-lined collar and hand-stitched sponsor patches from defunct fictional brands, layered over a ribbed charcoal compression top, high-waisted wide-leg mechanic trousers in dark indigo canvas with reinforced knee panels and dangling carabiner key rings clipped to a canvas utility belt slung low on one hip, hands wrapped in fraying ivory hand-wrap tape extending past the wrists, and chunky platform steel-toe boots in scuffed matte black with thick ridged rubber soles and asymmetric buckle straps, showing a poised calm stance with weight shifted to one leg and arms relaxed at her sides, designed where 3D is only the base structure, maintaining highly stylized elongated lanky proportions with sharp chiseled skeletal structures and exaggerated long fingers. [gear_module]: On the right side, an array of close-up callout square panels highlighting macro texture painting details of the cracked aged oxblood leather grain and frayed wool collar fibers of her bomber jacket, blueprint-style vector schematic line drawings of her custom titanium steering wheel with thumb-trigger nitrous activation and ergonomic suede grip wrapping, and technical lists detailing her modified 1997 coupe specifications including sequential twin-turbo inline six, hydraulic handbrake integration, roll cage geometry, and a cracked rearview mirror she refuses to replace for superstitious reasons. [expression_module]: Running along the bottom quadrant, a perfectly aligned horizontal row of 5 isolated headshot expressions displaying Neutral with half-lidded confidence, Smirk with one corner of her mouth pulled into a sly knowing grin, Focused with narrowed eyes and clenched jaw studying a road ahead, Shocked with widened asymmetrical eyes and parted lips, and Aggressive with bared teeth and deep furrowed brow veins visible at the temple, all featuring highly stylized 2.5D planar anatomy with volumetric chiseled shadow planes, soft textured loc volumes with golden bead glints rendered as flat pigment shapes, and vibrating grease-pencil outlines with heavy line-weight variation. [style_tags]: hand-painted 2.5D, texture painting, 3D is only the base structure, vibrating grease-pencil outlines, Arcane style painterly textures, thick oil-painterly brushstrokes with visible impasto canvas grain, non-action character study, deep ambient occlusion shadow planes, stepped cel-shading, printer halftone dot textures in transitional shadows, graphic ink slashes on clothing folds, matte sophisticated color blocking, masterwork model pack.
gpt-image-2.5-flare · medium · 1536x864 · job J57Result: All required views: front, side and back, a swinging action pose, the face with and without the mask, glove, belt and boot details, the grapple device and a swing sequence. Colours as given, plus a small name and stats block.
Works because. Bracket slots for identity, outfit, palette and action, with a fixed sheet: front, side and back views, a three-quarter action pose, a face close-up, equipment details.
Carries the effect. “same face and body structure in every pose, clear costume continuity”, plus rules against copying known characters.
To reuse. Fill every slot. Empty slots get generic choices.
Sent for the image: Slots filled: hero; Vela Kade, night-shift bridge rescuer; 28, human, tall and lean; calm, dry-humoured protector; magnetic grappling cables fired from wrist devices; a teal and charcoal padded rescue suit with orange reflective bands, segmented shoulder plates and a short hooded cape; teal, charcoal, signal orange; SWINGING.
Prompt, verbatim · 2,487 characters · 358 words
Show the full prose prompt
Create a highly detailed full-color character design sheet in 16:9 horizontal format for an original [CHARACTER TYPE / HERO / CREATURE / VILLAIN].
STYLE:
stylized cinematic character design,
high-quality animated feature look,
clean readable shapes,
strong silhouette,
premium concept art presentation,
bold graphic color palette,
comic-book inspired energy,
polished but production-ready design,
clear anatomy and costume readability.
CHARACTER IDENTITY:
[CHARACTER NAME / ROLE]
[AGE / SPECIES / BODY TYPE]
[PERSONALITY ARCHETYPE]
[POWER / SKILL / SPECIAL EQUIPMENT]
MAIN DESIGN:
The character should have a distinctive, memorable silhouette.
Use a clear costume language with recognizable shapes, strong color blocking, and functional details.
The design must feel original, not based on any existing franchise character.
No copyrighted logos, no recognizable existing superhero symbols, no direct imitation of known characters.
OUTFIT / ARMOR:
[DESCRIBE COSTUME OR ARMOR]
Include practical design details:
gloves / wrist devices
boots / shoes
belt gear
armor plates
fabric folds
glowing elements if needed
utility tools or weapons if needed
COLOR PALETTE:
[MAIN COLOR]
[SECONDARY COLOR]
[ACCENT COLOR]
Use a bold cinematic palette with strong contrast.
The colors should be clear enough for animation and video generation consistency.
CHARACTER SHEET LAYOUT:
Show the same character in multiple views on one clean sheet:
front view full body
side view full body
back view full body
three-quarter action pose
close-up face / mask expression
hand / glove / equipment detail
special ability or weapon detail
POSES:
Use confident readable poses.
The action pose should show the character’s main movement style:
[RUNNING / JUMPING / FLYING / SWINGING / FIGHTING / CASTING POWER / USING EQUIPMENT]
EQUIPMENT / POWER DETAIL:
Show how the character’s signature equipment or power works.
Example:
magnetic grappling cables,
energy gauntlets,
kinetic boots,
utility belt,
glowing power core,
mechanical wings,
elemental weapon,
or custom ability system.
BACKGROUND:
clean light neutral background,
minimal graphic design,
no complex environment,
no text-heavy poster design,
small visual notes allowed only if clean and readable.
QUALITY:
high detail,
sharp clean rendering,
consistent proportions across all views,
same face and body structure in every pose,
clear costume continuity,
production-ready character sheet,
suitable as a reference image for storyboard and AI video generation.
gpt-image-2.5-sunburst · high · 1024x1536 · job J58Result: The Chinese labels I checked are correct. 3 panels with glyphs and constellations, the same woman in all three. Not asked for: an English footer (“SAME SOUL DIFFERENT WAVES / THREE WATERS, ONE HEART”) and small Chinese captions.
Works because. Three panels, each with "identity": "same young woman model reimagined as zodiac character", then pose, outfit and background per sign. The constraints repeat “same woman as the base character”.
Carries the effect. One stable feature (long dark hair in a ponytail in all three) stays constant while outfit and palette change. A shared feature helps the eye read “same person”.
Watch out. Without an input image, “same woman” only means consistent within this one image. For a real person, send the photo.
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 3,786 characters · 329 words
Show the full JSON prompt
Line breaks added for reading. Copy gives the original single line.
{
"type": "Chinese zodiac-style character infographic poster",
"subject": "twelve zodiac character list, water signs edition",
"language": "Traditional Chinese",
"format": "vertical poster",
"style": {
"overall": "elegant anime-inspired character catalog with editorial infographic layout",
"rendering": "soft polished digital illustration, pastel gradients, delicate sparkles, ornamental border design",
"mood": "dreamy, celestial, refined, feminine, aquatic"
},
"canvas": {
"aspect_ratio": "2:3",
"background": "very light pearl white with pale blue-lavender tint, subtle texture, thin decorative frame with filigree corners and tiny stars"
},
"header": {
"title": "{argument name=\"headline text\" default=\"十二星座角色清單|水象星座\"}",
"subtitle": "感受・直覺・共鳴",
"icons": [
"small stars",
"water droplet emblem in top right",
"curled cloud-like line art in top left"
]
},
"layout": {
"sections_count": 3,
"sections": [
{
"title": "巨蟹座 Cancer",
"position": "top panel",
"theme_color": "powder blue",
"zodiac_symbol": "Cancer glyph inside circle at left",
"constellation": "Cancer constellation at upper right",
"count": 6,
"labels": [
"元素:水",
"概念:情感守護者,把人放在心上",
"性格:溫柔、敏感、顧家",
"行動原則:先確認感受,再保護重要的人",
"戀愛傾向:慢慢靠近,越熟越黏",
"人際怪癖:嘴上說沒事,實際會記很久"
],
"character": {
"identity": "same young woman model reimagined as zodiac character",
"pose": "half-body portrait, facing forward, arms gently wrapped around a large seashell pillow",
"hair": "long dark hair in a low ponytail",
"outfit": "light blue celestial slip dress with lace trim and sheer cardigan embroidered with stars and moons",
"accessories": "minimal jewelry",
"background": "soft blue night sky with crescent moon, seashell, sparkling stars, stylized ocean wave and tiny water droplets"
}
},
{
"title": "天蠍座 Scorpio",
"position": "middle panel",
"theme_color": "deep violet",
"zodiac_symbol": "Scorpio glyph inside circle at left",
"constellation": "Scorpio constellation at upper right",
"count": 6,
"labels": [
"元素:水",
"概念:深海偵察者,情緒有深度",
"性格:專注、神秘、意志強",
"行動原則:先觀察,再一擊到位",
"戀愛傾向:愛得深,重忠誠與獨占感",
"人際怪癖:越在乎越不說,會偷偷試探"
],
"character": {
"identity": "same young woman model reimagined as zodiac character",
"pose": "half-body portrait, one hand near chin in a composed, enigmatic gesture",
"hair": "long dark ponytail",
"outfit": "black semi-sheer dress with gothic details and a dark plum off-shoulder shawl",
"accessories": "dangling earrings and layered necklace",
"background": "dark purple celestial sea scene with crescent moon, bubbles, stars, and curling misty water shapes"
}
},
{
"title": "雙魚座 Pisces",
"position": "bottom panel",
"theme_color": "lavender",
"zodiac_symbol": "Pisces glyph inside circle at left",
"constellation": "Pisces constellation at upper right",
"count": 6,
"labels": [
"元素:水",
"概念:夢境共感者,靠直覺導航",
"性格:浪漫、柔軟、有想像力",
"行動原則:先感受,再順流找答案",
"戀愛傾向:容易心動,渴望靈魂陪伴",
"人際怪癖:常把別人的情緒也一起感受"
],
"character": {
"identity": "same young woman model reimagined as zodiac character",
"pose": "half-body portrait, one hand lifted as if balancing floating bubbles, other hand resting lightly at chest",
"hair": "long dark ponytail with a pale flower hair ornament",
"outfit": "translucent lavender fantasy dress with soft draped sleeves and shimmering fabric",
"accessories": "delicate earrings and necklace",
"background": "pale lilac underwater-celestial blend with bubbles, sparkles, and flowing translucent wave forms"
}
}
],
"dividers": "three horizontal framed panels with thin ornamental borders"
},
"footer": {
"center_icon": "small blue seashell emblem",
"decorations": [
"tiny stars",
"fine scrollwork"
]
},
"constraints": [
"all three zodiac entries must use the same woman as the base character with different styling, clothing, pose, and mood",
"text should be clean, editorial, and readable",
"each panel should clearly separate illustration area on the left and text block on the right",
"maintain cohesive water-element theme across all 3 signs",
"do not include the other nine zodiac signs in this image"
]
}
46
Person with a plush alter ego
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1024x1536 · job J23 · input: J01Result: Head to toe with a tree-spirit plush; the outfit was completed with jeans and sneakers. No text or logos.
Works because. The mascot is derived from the person (“inspired by the subject's mood, facial impression, styling”), with palette cues from hair, skin and clothing. Both must be visible head to toe.
Carries the effect. “If the original image only shows part of the outfit, intelligently complete the full look”. Needed here, because the input is waist-up.
Watch out. “Automatically choose a creature concept” gives a different creature each run. Good for exploring; name the creature once you like one.
Prompt, verbatim · 2,278 characters · 326 words
Show the full prose prompt
Use the uploaded portrait as the identity reference and preserve the person's recognizable facial features, hairstyle, skin tone, expression, fashion sense, and overall presence. Create a premium full-body portrait of the same person alongside a large custom-designed plush companion that feels like their mascot alter ego. The plush should be inspired by the subject's mood, facial impression, styling, posture, and overall energy rather than being a generic animal or mascot. Automatically choose a creature concept that best matches the person's unique vibe, avoiding predictable or stereotype-based selections. The mascot must clearly be an oversized plush toy with soft fuzzy fabrics, rounded shapes, detailed stitching, premium textures, and a collectible designer-toy aesthetic. Its design, expression, silhouette, and proportions should subtly reflect the person's character and visual identity. Build a harmonious color palette using cues from the subject's hair, skin tone, clothing, and atmosphere so the person, mascot, and scene feel naturally connected. Show both the person and plush fully visible from head to toe, including shoes and all parts of the mascot, with balanced framing and comfortable spacing. Choose a natural interaction that suits the subject, such as standing beside, sitting with, leaning on, lightly hugging, or casually engaging with the plush companion. Keep the person's expression relaxed, warm, and authentic with a subtle smile or calm gaze, avoiding stiff poses or mannequin-like appearances. If the original image only shows part of the outfit, intelligently complete the full look in a believable and stylish way. Place the scene in a clean, aesthetically pleasing environment such as a minimalist studio, cozy lifestyle setting, or refined editorial backdrop that complements both the person and mascot without distractions. The final image should feel charming, cozy, stylish, emotionally engaging, visually cohesive, and suitable for a high-end character campaign or social-media editorial. Avoid cropped bodies, hidden shoes, incomplete mascot visibility, generic animal choices, real animals, horror elements, cheap toy aesthetics, awkward poses, cluttered backgrounds, distorted anatomy, extra limbs, text, logos, or watermarks.
47
Photo to 3D chibi figure
ProseEdit: needs an input imageSuggested model: Sunburst
Inputgpt-image-2.5-sunburst · high · 1024x1536 · job J24 · input: J01Result: Stylised realism, not chibi proportions: the prompt's last section asks for “stylized realism” and that won. “Pixel & Coffee Club” is on the top, large rather than a tiny patch. Not asked for: many invented notes and a cat.
Works because. Scene, outfit, floating props and mood are separate lists. The props tell a story about the person (editing timelines, sticky notes, low productivity meters).
Watch out. Chibi is heavy stylization, and identity survives it worst. For figurine or 3D styles, first make a six-view sheet of the person (front, both three-quarters, both profiles, back), check it, then render from the sheet.
Check. “Tiny embroidered logo reading "Pixel & Coffee Club"” asks for quoted text on a tiny patch. Read it.
Prompt, verbatim · 2,535 characters · 346 words
Show the full prose prompt
Use my uploaded photo as the sole facial reference. Preserve my exact identity, facial proportions, eye shape, skin tone, lip shape, nose structure, and all distinctive features with exceptional accuracy and no facial drift.
Create a highly detailed cinematic 3D chibi-style character portrait inspired by top-tier animated films, luxury collectible figurines, and contemporary designer toy aesthetics. The character represents a brilliant AI-powered digital creator running on pure creativity and almost no sleep after a marathon night of editing and designing.
Scene:
Early morning inside a cozy creative studio. She sits drowsily on the edge of a weathered wooden workbench, legs gently swinging, looking exhausted yet irresistibly charming. Her fluffy hair is gathered into a loose messy bun with flyaway strands escaping in every direction. Sleepy half-open eyes, subtle under-eye shadows, a tiny rosy nose, and an expression that perfectly captures "I've been awake for way too long but I still have ideas."
Outfit:
Oversized soft-cream pajama set
Tiny embroidered logo reading "Pixel & Coffee Club"
Relaxed knitted cardigan casually slipping from one shoulder
Plush cloud-shaped slippers
Minimal pearl stud earrings
Cozy oversized sleeves partially covering her hands
Environment:
Warm sunrise light fills the room through large studio windows. A steaming mug of coffee rests beside her. The background is softly blurred, filled with artistic clutter, creative tools, and subtle storytelling details that suggest endless projects in progress.
Personality Elements Floating Around Her:
Mini holographic editing timelines
Tiny floating storyboard frames
Digital sketch concepts
Glowing location pins and travel stickers
Sticky notes filled with random ideas
Compact laptop displaying an unfinished creative project
Pencil sketches and concept doodles
Floating play-button icons
Small animated stars, clouds, and creative symbols
Tiny productivity meters running dangerously low
Expression & Mood:
Brilliant but overworked creator energy
Lovably exhausted
Chronic "one last revision" mindset
Quietly chaotic genius
Soft humor mixed with creative burnout
Dreamy, relatable, and emotionally expressive
Art Style:
Ultra-premium stylized realism, luxury designer collectible quality, highly detailed hair fibers, realistic fabric textures, soft skin rendering, cinematic depth of field, subtle emotional storytelling, cozy lifestyle aesthetic, handcrafted figurine finish, social-media-worthy presentation.
gpt-image-2.5-sunburst · high · 1024x1536 · job J59Result: Figure, sculpt and render match each other, with the nameplate “INES HALLORAN CARTOGRAPHER”. ZBrush on the monitor comes from the prompt (“similar to ZBrush”). The braid is long, not short. Not asked for: a “BETTER CHARACTERS A BRIGHTER WORLD” poster, an “Art Builds a Kinder World” cup and a “THE DRY LANDS” map.
Works because. The same character appears three times in one frame: a resin figure in front, a gray sculpt on the left monitor, a coloured render on the right. Each is told to match the others.
To reuse. Fill the brackets. With an image of the character, attach it and the prompt becomes an edit.
Sent for the image: Slots filled: Ines Halloran, an original desert cartographer character; freckled light-brown skin, a narrow face and grey-green eyes; a short copper braid; curious, quietly confident; a sand-coloured canvas field coat over a rust-red vest, with a leather map satchel and tall laced boots; standing with one foot on a small rock, unrolling a paper map with both hands; a brass compass and a rolled map at the base.
Prompt, verbatim · 2,689 characters · 388 words
Show the full prose prompt
Photorealistic high-quality studio photo of a modern digital art workspace, showing the concept of “from 3D virtual character to real collectible figure.”
In the foreground, a highly realistic collectible figurine of [Character Name / Character Identity] is placed on a round wooden display stand. The character has [facial features / appearance], [hairstyle], and a [expression / personality vibe]. The figure is wearing [outfit / costume]. The overall design is refined, premium, and instantly recognizable. The figurine should have realistic collectible statue quality, with subtle resin/sculpture material feel, while still looking highly believable and visually realistic.
The pose is [character pose], natural, stable, elegant, and display-worthy. Shot from a low-angle close-up perspective with slight wide-angle distortion, vertical composition, emphasizing the full figure, clothing structure, leg lines, and pose.
In the background, there is a professional 3D character design workstation with two large curved monitors. Both monitors must show the exact same character as the foreground figurine — same face, same hairstyle, same outfit, same pose, and same overall vibe — clearly expressing the idea of turning a digital 3D character into a real physical figure.
The left monitor shows a gray sculpt / clay model view in a professional 3D sculpting software interface, similar to ZBrush. The gray model must match the foreground figure exactly in character design, pose, outfit structure, and facial identity.
The right monitor shows the fully rendered colored version of the same character, also matching the foreground figure exactly in face, hairstyle, outfit, pose, and temperament. Together, the two monitors reinforce the workflow of “digital character design → physical collectible statue.”
On the desk are a keyboard, mouse, monitor arms, drawing tablet, stylus, and other 3D modeling tools. The workspace is clean, professional, and visually premium. Optional extra elements: [weapon / accessories / theme props / IP-style design details].
Lighting is a mix of soft studio lighting and indoor workspace lighting. The foreground figurine is evenly lit with clear facial and material detail, while the monitors emit cool-toned tech light. Overall mood is realistic, clean, premium, slightly shallow depth of field, ultra-detailed, emphasizing the collectible figure quality, professional 3D design studio atmosphere, and the visual concept of “from digital model to real figure.”
photorealistic, ultra detailed, cinematic studio lighting, realistic figurine, collectible statue, 3D character design studio, from digital model to real figure, vertical composition
gpt-image-2.5-sunburst · high · 1248x1760 · job J60Result: All 18 sections with their titles. 11 expressions, 9 poses, 5 + 5 turnaround views, 4 mockups, 6 icons and 5 swatches are correct; section 04 has 15 sketches instead of 12. The small text is legible. Not asked for: “Good Tea Brighter Days”, “TEA PEOPLE A BRIGHTER YOU”, “A CUP OF GOODNESS A HAPPIER TOMORROW”, “SMALL LEAF BIGGER HAPPINESS”.
Works because. 18 sections in a 3x6 grid, each with element counts: 12 concept sketches, 11 expressions, 9 poses, 5 turnaround views.
Watch out. I expected most sections to come out as decoration. Sunburst rendered all 18 legibly, but a dense grid is where count errors hide: check each section's numbers. For assets you will use, generate the parts you need (expressions, turnaround) as separate images.
Also in the repo. A near-duplicate 18-panel version by @Colin_Leeee.
Sent for the image: Template slots filled with their defaults (MUYANG TEA, tea shop, yellow and green, a Shiba Inu in a green apron).
Prompt, verbatim · 3,407 characters · 384 words
Show the full JSON prompt
{
"type": "18-section complete brand identity and mascot design sheet",
"brand": {
"name": "{argument name=\"brand name\" default=\"MUYANG TEA\"}",
"industry": "{argument name=\"industry\" default=\"tea shop\"}",
"colors": ["{argument name=\"primary color\" default=\"yellow\"}", "{argument name=\"secondary color\" default=\"green\"}", "white", "brown", "dark green"]
},
"subject": "{argument name=\"character description\" default=\"3D rendered cute Shiba Inu mascot wearing a green apron\"}",
"layout": {
"grid": "3-column by 6-row grid layout",
"sections": [
{
"title": "01 BRAND DNA ANALYSIS",
"elements": ["brand logo", "5 color swatches", "6 brand icons", "target audience charts"]
},
{
"title": "02 CONCEPT MOODBOARD",
"elements": ["5 reference photos", "4 mood icons", "design concept equation"]
},
{
"title": "03 FORM STUDY",
"elements": ["4 logo anatomy icons", "4 design evolution steps", "4 character silhouettes"]
},
{
"title": "04 CONCEPT EXPLORATION",
"elements": ["12 line-art character concept sketches"]
},
{
"title": "05 REFINED LINE ART",
"elements": ["3 rows of front and side view line art with proportion guides"]
},
{
"title": "06 DETAIL REFINEMENT",
"elements": ["2 full-body renders with annotation labels", "4 circular close-up views"]
},
{
"title": "07 EXPRESSION SHEET",
"elements": ["11 3D rendered facial expressions"]
},
{
"title": "08 POSE LIBRARY",
"elements": ["9 full-body 3D rendered character poses"]
},
{
"title": "09 TURNAROUND VIEW",
"elements": ["5 full-body 3D renders from multiple angles", "5 matching line-art views"]
},
{
"title": "10 COLOR DEVELOPMENT",
"elements": ["5 rows of 5-color palette options", "color psychology explanations"]
},
{
"title": "11 MATERIAL SPECIFICATION",
"elements": ["5 surface texture swatches", "material property sliders", "4 manufacturing process icons"]
},
{
"title": "12 COLOR APPLICATION",
"elements": ["4 color scheme variant renders", "2 light and dark mode renders", "4 contrast rating indicators"]
},
{
"title": "13 CONSTRUCTION GUIDE",
"elements": ["2 line-art technical diagrams for geometry and grid system"]
},
{
"title": "14 DESIGN SYSTEM RULES",
"elements": ["minimum size icons", "clear space diagram", "4 correct and incorrect usage examples"]
},
{
"title": "15 ASSET VARIANTS",
"elements": ["3 scaled size variants", "3 line-art style variants", "3 simplified flat icon heads"]
},
{
"title": "16 DIGITAL APPLICATIONS",
"elements": ["1 app icon design", "2 social media avatar versions", "UI component elements", "3-frame animation cycle"]
},
{
"title": "17 PHYSICAL APPLICATIONS",
"elements": ["plush toy product mockup", "product packaging mockup", "branded merchandise mockup", "retail storefront mockup"]
},
{
"title": "18 FINAL RENDERING",
"elements": ["large high-resolution 3D mascot render holding tea cup", "finalized logo", "deliverable file format list"]
}
]
}
}
JSON vs prose
JSON by itself does not improve GPT Image output. OpenAI's 2.5 guide says prose, JSON-like structures, instructions and tags “can all express the same intent”, and no controlled 2026 test on Flare or Sunburst shows a format effect; the side-by-sides that exist (Neurl Creators, Chris Green) find little difference against equally detailed prose. What helps is structure: it makes you fill every slot, keeps “change” apart from “preserve”, and lets you swap one field at a time. JSON is one way to get structure; labeled sections are another.
Most viral “JSON wins” compare a detailed JSON prompt against a short prose one. The gain comes from the extra information. JSON prompts written by an LLM also tend to run long: the auto-generated JSON prompts in the YouArt repo are about 35% longer than prose for the same result.
Text-heavy layouts. Both count elements and quote every string; the labeled prose sections do the same job.
A controlled pair
The ice cream JSON (#7) sent verbatim, and my prose rewrite of the same fields in the same order, sent to the same model at the same settings. The duplicated label (“Chocolate ice cream with brownies” twice) stays in both, to show that the format does not fix the content. One sample each, so this shows the kind of difference, not a measurement.
JSON (verbatim) · gpt-image-2.5-sunburst · high · 1152x1536 · job J06Result: All four labels as written, including the duplicated brownie label on the vanilla side.Prose (my rewrite) · gpt-image-2.5-sunburst · high · 1152x1536 · job J07Result: Nearly the same image as J06: same four labels, same duplicate. The almonds on top are whole instead of chopped.
My prose rewrite of the ice cream JSON (job J07)
Show the prose version
Photorealistic commercial product render, 3:4 vertical. A single ice cream bar on a wooden stick, centered, upright and front-facing, seen from a slightly elevated angle, with ingredient labels and curved arrows arranged around it. Background: a smooth matte warm golden-yellow gradient, evenly lit, with a subtle vignette darker toward the edges. Studio lighting: a soft frontal key light that brings out the gloss of the chocolate, a balanced fill that keeps texture detail, visible specular highlights on the melted chocolate coating, and a soft shadow beneath the stick. The bar is a rounded rectangle with visible embedded inclusions. Its top is coated in glossy, thick, fluid melted milk chocolate with several drips running down, rounded drip edges pulled by gravity and irregular almond pieces set into the coating. The left interior is dense, creamy chocolate ice cream with small dark brownie chunks spread evenly, matte against the glossy coating. The right interior is smooth, creamy vanilla ice cream with light beige caramel chunks with rounded edges. The stick is light natural wood, smooth with subtle grain, flat with rounded edges, fully visible below the bar. Ingredient callouts use thick curved dark-brown arrows and dark-brown clean sans-serif text, balanced around the product: top-left "Chocolate with almonds" with whole almonds and small chocolate squares; left-middle "Chocolate ice cream with brownies" with brownie chunks; right-middle "Chocolate ice cream with brownies" with brownie chunks; bottom-right "Vanilla ice cream with caramel pieces" with caramel cubes and white vanilla pieces. Colours: milk chocolate brown, golden yellow and cream white, with dark brownie brown, light caramel orange and almond beige as secondary colours. Extreme micro-detail, high texture realism, no noise, moderate depth of field with the whole product in focus. Luxury dessert advertising, hyper-realistic food photography, clean studio composition. 8K.
When to use which
Task
Form
Single portrait or product hero
Prose, in the order scene, subject, details, constraints
Infographic, UI, character sheet, grid, poster with text
JSON or labeled sections
Campaign series, catalogue variants
A template with fixed fields (palette, light, lens, framing) and variable fields (subject, action, place, copy)
Edits, identity work
An edit contract: core rule, preserve, change, negative; repeated every turn
Agent or script pipelines
JSON you can validate and vary in code
Converting JSON to prose
Keep the key order; it is usually scene, subject, details, style.
Turn each object into one sentence and each array into a list inside that sentence.
Drop keys that are really API settings (resolution, aspect_ratio, quality, params) and set them as parameters.
gpt-image-2.5-flare · medium · 1536x1536 · job J61Result: All 16 poses match their descriptions, in order, with blurred faces and the same outfit in every cell.
What it is. The same content twice: a JSON spec with 16 numbered pose descriptions, and a prompt field in prose that describes the sheet in general terms.
Which part does the work. The prose field alone gets you a generic pose sheet. The numbered pose list is what pins down each cell. The JSON adds information, not format.
To reuse. For a pose chart, keep the pose list and the camera line (straight-on, consistent distance).
Sent for the image: Template slots filled with their defaults.
Prompt, verbatim · 3,554 characters · 410 words
Show the full JSON prompt
Line breaks added for reading. Copy gives the original single line.
{
"type": "pose reference sheet",
"subject": {
"category": "female dancer fitness model",
"age_appearance": "young adult",
"build": "slim athletic",
"hair": {
"color": "dark brown",
"style": "high ponytail"
},
"outfit": {
"top": "light gray or white sports bra crop top",
"bottom": "baggy light gray sweatpants",
"shoes": "white sneakers"
},
"face": "softly blurred or de-emphasized facial features"
},
"style": {
"image_type": "studio dance pose chart",
"background": "clean seamless white background",
"lighting": "bright even studio lighting with minimal shadows",
"color_palette": "neutral whites and light grays",
"camera": "full-body framing, straight-on view, consistent distance",
"rendering": "photorealistic"
},
"layout": {
"grid": {
"rows": 4,
"columns": 4,
"count": 16,
"border": "thin black dividers between cells"
},
"numbering": {
"count": 16,
"labels": [
"1",
"2",
"3",
"4",
"5",
"6",
"7",
"8",
"9",
"10",
"11",
"12",
"13",
"14",
"15",
"16"
],
"position": "top-left corner of each panel"
},
"sections": [
{
"title": "row 1",
"position": "top",
"count": 4,
"labels": [
"1 side lunge with one arm extended straight sideways and the other bent near chest",
"2 low floor pose leaning on one hand with one knee down and opposite arm arched upward",
"3 wide squat facing front with both arms opened in angular dance position",
"4 standing balance on one leg with opposite knee lifted and forearms crossed near chest"
]
},
{
"title": "row 2",
"position": "upper-middle",
"count": 4,
"labels": [
"5 deep backbend in wide stance with torso arched and one arm curved overhead",
"6 wide squat with one hand behind head and the other arm pointing outward",
"7 kneeling side stretch with one hand on floor and opposite arm reaching straight up",
"8 standing arabesque-style extension with torso tilted forward and one leg lifted high behind/sideways"
]
},
{
"title": "row 3",
"position": "lower-middle",
"count": 4,
"labels": [
"9 wide squat with torso tilted left, one arm curved overhead and one arm extended low",
"10 front-facing wide squat with both arms stretched diagonally in opposite directions",
"11 relaxed standing pose with legs apart and both forearms crossing in front of torso",
"12 floor recline supported on one hand and one knee, torso leaning back with bent legs"
]
},
{
"title": "row 4",
"position": "bottom",
"count": 4,
"labels": [
"13 small jump or lifted balance with one knee raised and one arm bent upward",
"14 low crouch squat with one hand reaching toward floor and other arm extended sideways",
"15 dramatic side backbend in wide stance with hair swinging and one arm curved overhead",
"16 powerful wide squat with one hand at chest and the other lowered to the side"
]
}
],
"overall_composition": "all 16 poses shown as separate panels in a uniform contact sheet"
},
"prompt": "Create a clean studio contact sheet of {argument name=\"pose count\" default=\"16\"} full-body dance or combat-reference poses featuring a {argument name=\"subject type\" default=\"young athletic woman\"} in a {argument name=\"outfit\" default=\"light gray sports bra, loose gray sweatpants, and white sneakers\"}. Use a seamless {argument name=\"background color\" default=\"white\"} background, bright even lighting, and a consistent straight-on camera. Arrange the poses in a 4x4 grid with thin black panel lines and small black numbers 1 through 16 in the top-left of each cell. The poses should mix standing, squatting, kneeling, floorwork, balance, kick-extension, backbend, and angular arm positions suitable for a dance sheet chart or combat movement reference. Keep the styling photorealistic, crisp, minimal, and instructional, with consistent wardrobe and hair across all panels."
}
Flare vs Sunburst
The skincare prompt (#9), sent as written to both models: Sunburst at high, Flare at medium (the cheap draft setting). One sample each. Look at the label text and at anything the prompt did not ask for.
Sunburst · gpt-image-2.5-sunburst · high · 1024x1280 · job J08Result: Label text correct. Not asked for: a line under the logo (“NATURAL BEAUTY FOR A BRIGHTER YOU”) and two slogans on the set (“NATURE NOURISHES A BRIGHTER YOU”, “PURE PLANTS BRIGHTER SKIN”).Flare · gpt-image-2.5-flare · medium · 1024x1280 · job J09Result: Label text correct, no added copy, a simpler set than J08.
Both get every label string right. The difference is in what was not asked for: Sunburst added a line under the logo and two slogans on the set, Flare added nothing and built a simpler set. At medium, Flare used about a quarter of Sunburst's output tokens (378 vs 1,510).
Added copy across the page
The later images (J25-J61) repeat the pattern. Of the 22 Sunburst images made from repo prompts in that set, 12 added text nobody asked for, and in 8 of them it is a slogan about brighter or better days: “A BRIGHTER KIND OF TOMORROW” (J33), “GOOD PEOPLE BETTER DAYS” (J47), “HEROES PROTECT A BRIGHTER TOMORROW” (J49), “Good Tea Brighter Days” (J60). Flare added copy in 2 of 11 (J31, J55). I did not test a fix; the usual one is to end the prompt with “no text other than the quoted strings”.
Output tokens in that set: Flare at medium used 280-628 per image, Sunburst at high 1,078-2,511.
Moderation: Flare was stricter in this run
The window-light portrait (#2) was meant to be the second pair. On 2026-09-26 both models refused it twice as written. Sunburst accepted a minimally edited version (lips closed, a modest scoop neck, the cardigan on the shoulders, jeans instead of shorts). Flare refused that exact version and 3 further rewrites, so there is no Flare image; Sunburst's result is on the card. One prompt is a small sample, and Sunburst's filter was not consistent either: it refused one of the stricter rewrites. In this run, Flare's output moderation was stricter than Sunburst's.
Later in the same run, Sunburst also refused the anatomy plate (#34) twice as written, an edit of a fully clothed fictional woman into a textbook figure. Two minimal rewrites were refused as well; the third, which drops two muscle labels and the organ silhouettes, passed. The diff is on the card. Every refusal on this page was output-stage moderation, category sexual, and every one involved a woman's body, skin or clothing. All 11 later Flare jobs passed on the first try, but none of them touched on those.
What to expect from other testers: in single shots the two are close; Flare gives cleaner skin and a warmer grade, Sunburst keeps more pores and handles edits and small text better.
Editing contract
An edit prompt is a contract: what to change, what to keep, and what must stay physically consistent. The edits of a person on this page all start from the same source image, a fictional person generated from my own prompt (J01). Each edit starts from the original, not from the previous edit. The anatomy plate needs a full-body photo, so its input is a full-body version of J01 made in one step (J25), shown as its thumbnail.
Source image · gpt-image-2.5-sunburst · high · 1152x1536 · job J01Result: Matches the prompt. The chalkboard has faint marks but no readable text.
My prompt for the source image (job J01)
Show the prompt
Photorealistic smartphone photo of a fictional adult woman in her early 30s sitting at a wooden café table by a large window, looking straight at the camera with a relaxed, slight smile. Shoulder-length dark brown hair with a few loose strands, light olive skin with visible pores and a few faint freckles, no heavy makeup. She wears a plain oatmeal crew-neck knit sweater and holds a white ceramic mug with both hands on the table. Soft overcast daylight from the window on the left, gentle shadows on the right side of her face. Background: a softly blurred café interior with plants and a chalkboard with no readable text. Eye-level, 3:4 vertical, waist-up framing, natural colours, slight phone-camera noise. No text, no logo, no watermark. Both hands relaxed around the mug.
It describes the capture (smartphone, overcast window light from the left, slight noise), the imperfections to keep (pores, freckles, loose strands) and where the hands are.
Eight edits of one person
Identity reference in JSON · gpt-image-2.5-sunburst · high · 1152x1536 · job J16 · input: J01Result: Identity kept; headphones, hand pose and red object as asked. She smiles instead of the intense gaze the prompt asks for.Identity-locked expression · gpt-image-2.5-sunburst · high · 1024x1536 · job J17 · input: J01Result: Identity and the new expression are there. The hair is only a little longer than in the input, not the long hair the prompt asks for.Stone shards · gpt-image-2.5-sunburst · high · 1152x1536 · job J18 · input: J01Result: Body, hair and mug are stone shards. The face stays close to photographic skin with seams, not slate.Storybook split · gpt-image-2.5-sunburst · high · 1152x1536 · job J19 · input: J01Result: Photo and drawing match. The caption is the sample text copied as is, “little happy girl ♡”, although the prompt says to adapt or drop it when the subject is not a girl.Plush alter ego · gpt-image-2.5-sunburst · high · 1024x1536 · job J23 · input: J01Result: Head to toe with a tree-spirit plush; the outfit was completed with jeans and sneakers. No text or logos.“3D chibi” prompt · gpt-image-2.5-sunburst · high · 1024x1536 · job J24 · input: J01Result: Stylised realism, not chibi proportions: the prompt's last section asks for “stylized realism” and that won. “Pixel & Coffee Club” is on the top, large rather than a tiny patch. Not asked for: many invented notes and a cat.Pencil sketch, real mug · gpt-image-2.5-sunburst · high · 1024x1792 · job J54 · input: J01Result: The J01 likeness holds, with the face in graphite. The jacket is pencil too, not full colour as asked; hand and mug are photoreal. KAVA and K.A.V.A as written. The steam rises in front of her chin instead of covering the lower face. Not asked for: “SAME PEOPLE BRIGHTER DAYS” on a notebook and “Good Coffee Better People” on a bag.InputAnatomy plate · gpt-image-2.5-sunburst · high · 1024x1536 · job J50 · input: J25Result: An anatomy plate in the source pose, with the sweater and jeans as outlines. The Latin labels I checked are spelled correctly, with skull, spine, pelvis, hand and foot insets and the caption “Tabula Anatomica — Figura I / Studium Corporis Humani”. The face stays a photo instead of becoming a drawing. The scapula has no label.
The rules
Change one thing per turn, described physically (“replace the sweater with a dark gray tank top”), not by mood.
Repeat the full preserve list every turn. Identity, pose, framing, camera angle, light direction, background, clothing, labels. Shrinking it is one of the six common mistakes.
Ask for physics: new shadows, reflections and contact points for anything added.
Stop chaining after 2-4 rounds. Texture and identity degrade. Start a new call with the original, or the last accepted image, and the full description.
On a bad result, go back to the last accepted image, not forward into another fix.
Masks are soft. For pixel-exact regions, composite the approved edit into the original yourself.
Number the input images and give each a role (subject, style, clothing, background), and say which wins in a conflict. Identity references first; 1-4 clean ones are better than many.
Stylize after identity is stable. Anime, chibi and comic ink eat identity. For figurine or 3D styles, first make a six-view sheet of the person, check it, then render from it.
Two blocks to copy
An edit contract in JSON (abridged from a Visual AI Club post, 2026-09-24):
Edit contract, JSON
Show the contract
{
"CORE_RULE": "Use the supplied image as a locked photographic template. Change only the listed items.",
"PRESERVE": ["exact pose and body position", "camera angle, height, perspective and framing",
"background and environment", "lighting and shadows", "all clothing exactly as shown"],
"CHANGE": ["<one thing>"],
"PHOTOGRAPHY": {"style": "authentic candid smartphone photograph",
"skin": "natural pores and realistic imperfections", "retouching": "minimal"}
}
A preservation block for people, in prose:
Preservation block, prose
Show the block
Image 1 is the subject. Preserve her exact likeness: face shape, bone structure,
eye shape and eyelid fold, nose, jawline, skin tone, freckles and moles, hairline,
apparent age, expression and body proportions. Do not beautify, slim, smooth skin,
whiten teeth, enlarge eyes, or make her younger or more symmetrical.
Change only: [the one thing].
Keep unchanged: pose, framing, lighting direction, background.
What I left out
Prompts built around real brands, real products, real people or existing game characters, and rows themed on real sports tournaments.
Video prompts, which the repo includes alongside image prompts.
A political satire poster, and a prompt whose example carried another service's watermark.
Near-duplicates; where one exists, the card mentions it.
The repo's infographic and UI-mockup rows, except where they are storyboards, edits or character sheets. They were not on my list.
Style library: 103 restyle prompts from Visual AI Club
This is every ChatGPT image prompt on visualaiclub.com as of 2026-09-26: 103 styles with 705 example images. 94 of the styles turn a photo of people into something else. The prompts and images belong to their authors: each card names the author and the licence the author picked on the site (88 attribution, 15 CC0) and links to the source page. The site shows the first lines of most prompts and charges a copy credit for the rest. Visual AI Club also posts full prompts in Google Docs linked from its Telegram channel, and on X. Full prompts are shown here for 95 of the 103 styles: 85 from those Google Docs, 8 from X posts by Visual AI Club or the prompt's author, and 2 complete on the site. Each card links the source. For 4 more I found only a variant, with different placeholder names or a missing line; it is shown under the site's excerpt with the difference named. The other 4 show the site's excerpt verbatim and say how many lines are hidden; a line ending in … is cut short on the site too. The example images are the site's, not mine, and most include the original photo as a small inset.
I left out 97 of the 705 example images: 83 show real celebrities or athletes, and the rest are sexualised or contain rude text or smoking. For 10 styles that leaves no example here, and each card with a gap links to the full set on the site.
The styles are grouped by what they make, printable figures first, so that similar styles sit next to each other and you can see how the prompts differ. Each group starts with a strip of covers, one per style; click a cover to jump to its card. The figure-potential badge on each card is my rating by style type, not the site's.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 2,262 characters
Show the full prompt
Transform the uploaded person into an ultra-realistic Pixar-inspired 3D chibi character while preserving the exact facial identity, facial structure, hairstyle, skin tone, eye shape, smile, and overall likeness from the reference image.
FORMAT LOCK
Vertical 9:16 composition. Full-body character. Centered composition. Premium vacation artwork. Keep the character proportions consistent except for the oversized chibi head.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, skin tone, facial proportions, expression, and recognizable identity. Do not redesign the face or create a different person.
CHARACTER
Create a premium Pixar-inspired 3D chibi with a large adorable head, expressive eyes, rosy cheeks, soft cinematic facial shading, and shoulder-length black hair matching the reference image. The subject is relaxing on a tropical white-sand beach, sitting on a miniature wooden beach chair with crossed legs while holding a refreshing orange juice topped with an orange slice. She wears an oversized premium black T-shirt, white sneakers, silver hoop earrings, and has a relaxed, confident vacation expression.
COMPANION
Beside her sits an adorable French Bulldog wearing a matching premium black T-shirt, relaxing on its own miniature wooden beach chair with crossed paws.
BEACH DETAILS
Include a woven picnic basket, yellow rubber duck, seashells, starfish, detailed sandcastle, yellow-and-white beach umbrella, yellow inflatable swim ring, tropical fruits, and a small vintage camera naturally placed in the sand.
BACKGROUND
Crystal-clear turquoise ocean, gentle waves, bright blue sky, fluffy white clouds, distant tropical islands, vibrant summer atmosphere, warm golden sunlight, soft cinematic shadows, shallow depth of field, creamy bokeh, ultra-detailed sand textures.
STYLE
Ultra-realistic Pixar-inspired 3D rendering. Luxury vacation aesthetic. Premium color grading. HDR lighting. Soft global illumination. High-end animated film quality. Ultra-detailed 8K masterpiece.
NEGATIVE PROMPT
Avoid low quality, blur, extra fingers, extra limbs, duplicate characters, distorted face, bad anatomy, cropped body, plastic skin, oversaturated colors, text artifacts, watermark, logo, anime style, flat cartoon rendering, unrealistic lighting.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 2,212 characters
Show the full prompt
Transform the uploaded person into a heartwarming Pixar-inspired 3D chibi scene while preserving the exact facial identity, facial structure, hairstyle, skin tone, expression, outfit, colors, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Clean white seamless studio background. Bright centered composition. Premium greeting-card aesthetic. The giant text remains the central structure of the scene.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, skin tone, body proportions, outfit, shoes, and overall identity from the uploaded image. Do not redesign the face or change the clothing.
MAIN CHARACTER
Create a premium Pixar-style 3D chibi version of the uploaded person sitting comfortably on top of giant pastel 3D letters spelling "GOOD MORNING." The character should have a warm smile and relaxed pose while holding a ceramic coffee mug. The mug should remain simple with no visible text or logo.
MINI CHARACTERS
Surround the main character with multiple adorable miniature versions of the same person, each performing a different cheerful morning activity: pouring coffee, holding an encouraging sign, enjoying a coffee, reading a book, watering a small plant, hugging a pink heart, serving a cupcake, and raising a motivational chalkboard. Every mini character must preserve the same recognizable facial identity in a playful Pixar-style design.
DECORATION
Include cute coffee cups, latte art, tiny plants, tulips, a small orange kitten, wooden signs with positive messages, soft hearts, subtle doodles, and cozy decorative elements. Keep the scene balanced, clean, and uncluttered.
LIGHTING & STYLE
Soft natural studio lighting. Warm pastel color palette. Premium Pixar-inspired 3D character design. Expressive faces. High-quality textures. Cozy morning atmosphere. Minimal yet vibrant composition. Luxury animated-film aesthetic. Ultra-detailed 8K rendering.
NEGATIVE PROMPT
Avoid low quality, blurry details, distorted face, changed identity, altered outfit, extra limbs, duplicate main character, bad anatomy, plastic textures, cluttered composition, dark lighting, harsh shadows, logos, watermarks, text artifacts, or unrealistic 3D rendering.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 1,981 characters
Show the full prompt
Transform the uploaded person into a premium studio portrait featuring both a realistic human version and a Pixar-inspired 3D mini version of the same person.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Clean centered composition. Minimalist light-gray seamless studio background. Premium commercial photography.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial identity, facial structure, skin tone, hairstyle, glasses, tattoos, outfit, body proportions, and recognizable appearance. Both characters must clearly represent the same person.
REALISTIC CHARACTER
Place the realistic person behind the mini character with a relaxed, confident posture. Rest one hand naturally on the mini character’s oversized head. Maintain realistic skin texture, clothing details, and professional studio-quality lighting.
PIXAR MINI CHARACTER
Create a premium Pixar-inspired 3D mini version of the same person with a large expressive head, smaller body, bright eyes, soft facial features, and playful proportions. Preserve the same hairstyle, glasses, tattoos, clothing, and recognizable facial identity in a stylized form.
POSE & COMPOSITION
The Pixar mini character stands slightly closer to the camera while the realistic person stands behind. Both look directly at the camera with natural smiles and friendly interaction. The relationship should feel playful, warm, and visually balanced.
LIGHTING & STYLE
Soft professional studio lighting. Gentle shadows. Cinematic reflections. Ultra-detailed skin and fabric textures. Premium commercial photography. Pixar-quality 3D rendering. Luxury Instagram campaign aesthetic. Sharp focus. Ultra-realistic 8K quality.
NEGATIVE PROMPT
Avoid changing the person’s identity, hairstyle, glasses, tattoos, outfit, or proportions. Avoid blurry details, distorted anatomy, extra limbs, unrealistic expressions, plastic textures, cluttered backgrounds, text, logos, watermarks, or low-quality rendering.
7 examples are not shown here because they show real public figures. See all on the site ↗
You beside your Pixar-style chibi figurine
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 3,181 characters
Show the full prompt
Create a premium studio campaign portrait featuring the uploaded person standing beside a collectible Pixar-style chibi figurine version of themselves.
FORMAT LOCK
Vertical 4:5 composition. Full-body view. Both subjects fully visible from head to toe. Premium magazine-cover quality. Clean centered studio composition.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve exact facial structure, facial proportions, hairstyle, hair color, skin tone, expression, body proportions, clothing, accessories, jewelry, watch, shoes, and recognizable identity traits. The real person and the figurine must clearly represent the same individual.
STUDIO ENVIRONMENT
Medium neutral grey seamless backdrop with a subtle radial gradient, slightly brighter behind the subjects and darker toward the edges. Professional photography studio setup. Clean floor. No props. No distractions.
LIGHTING
Large softbox key light from upper-left. Gentle warm fill light from the right. Subtle rim light from behind. Soft premium studio shadows. Natural skin rendering. Both subjects must share identical lighting direction and shadow behavior.
REAL PERSON
Standing on the right side of the frame. Full height visible. Three-quarter stance toward camera. Relaxed confident pose. Weight on the back leg. Direct eye contact with the camera. Exact outfit and appearance from the reference image.
CHIBI FIGURINE
Standing on the left side, slightly in front of the real person.
Height reaches approximately the real person's chest level (about 60% of their height).
Balanced Pixar-style proportions. Head approximately 30% of total figure height. Body approximately 70% of total figure height. Full visible torso, arms, legs, and feet. Not a bobblehead. Not a Funko Pop. Not oversized head proportions.
Facial design should be a highly recognizable Pixar-style interpretation of the real person. Large expressive eyes with detailed irises and bright catchlights. Sculpted hairstyle matching the reference. Soft rounded cheeks. Warm friendly smile. Premium collectible figure quality.
WARDROBE MATCH
The figurine wears the exact same outfit, shoes, accessories, jewelry, watch, colors, patterns, and styling as the real person, adapted naturally to chibi proportions.
INTERACTION
The real person's right arm rests naturally on top of the figurine's head. Palm touching the hair. Relaxed elbow position. Comfortable natural interaction. Both subjects appear photographed together in the same studio at the same moment.
OUTPUT STYLE
Premium studio campaign photography. Photoreal human subject. High-end Pixar/Disney-style collectible figurine. Luxury advertising aesthetic. Realistic shadows. Seamless integration. Ultra-detailed materials. 8K quality. Fine film grain.
NEGATIVE PROMPT
Avoid bobblehead proportions. Avoid Funko Pop style. Avoid giant floating head. Avoid childlike proportions. Avoid mismatched outfit details. Avoid changing identity. Avoid different hairstyles. Avoid extra limbs. Avoid duplicate subjects. Avoid cartoon backgrounds. Avoid unrealistic shadows. Avoid low-detail figurines. Avoid text, logos, watermarks, or cluttered backgrounds.
Make the aspect ratio 3:4
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 1,779 characters
Show the full prompt
Create a premium stylized 3D animated avatar based on the uploaded reference image.
FORMAT LOCK
Vertical 3:4 composition. Medium portrait only (framed from approximately the waist or mid-torso upward). Character centered. Straight-on front-facing view. Symmetrical composition. Clean white seamless studio background.
IDENTITY LOCK
Preserve the exact facial identity, facial structure, hairstyle, skin tone, eyebrows, eyes, nose, lips, ears, facial proportions, clothing, colors, accessories, and recognizable appearance from the uploaded image.
POSE LOCK
Ignore the pose, camera angle, body orientation, hand position, and framing of the reference image.
Always recreate the character in the same default presentation:
• facing directly toward the camera
• shoulders squared and level
• upright relaxed posture
• arms naturally resting outside the frame or alongside the body
• friendly expressive facial expression
• centered composition
Never copy side profiles, tilted bodies, action poses, crossed arms, objects held in the hands, or unusual camera angles from the reference.
STYLE
Modern stylized animated 3D character, Pixar-inspired proportions, oversized expressive eyes, flawless smooth skin with subtle subsurface scattering, highly detailed hair strands, realistic fabric textures, soft cinematic studio lighting, premium global illumination, realistic soft shadows, shallow depth of field, ultra-clean minimalist aesthetic, ultra-sharp 8K, professional character design.
NEGATIVE PROMPT
Full body, side profile, three-quarter view, tilted body, dynamic pose, action pose, copied reference pose, hands near face, crossed arms, holding objects, cropped face, extra limbs, distorted anatomy, changed clothing, changed hairstyle, changed identity, text, watermark, logo.
8 examples are not shown here because they show real public figures. See all on the site ↗
Ghibli-style collectible 3D miniature
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 2,417 characters
Show the full prompt
Create a premium collectible 3D miniature version of the uploaded person, inspired by handcrafted Ghibli-style figurines.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 composition. Clean studio photography. Centered composition. Strong scale contrast between the miniature, giant chocolate, and human finger.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve facial structure, hairstyle, hair color, beard details, skin tone, facial proportions, and recognizable identity. The miniature must clearly resemble the uploaded person.
MINIATURE DESIGN
Cute handcrafted collectible figure with a slightly oversized head, small body, soft proportions, realistic sculpted face, crystal-glass eyes, sculpted eyelids, defined nose, soft smile, premium artisan toy quality.
POSE & STORY
The miniature is sitting on the floor, leaning slightly backward while hugging a giant chocolate bar with both hands.
The chocolate should be nearly as large as the miniature itself.
Several visible bite marks are already missing from the chocolate.
The miniature looks proud, stubborn, and slightly guilty, as if caught eating the chocolate but refusing to let go.
A giant realistic human finger gently touches the miniature's forehead, playfully trying to separate it from the chocolate.
CHOCOLATE DETAILS
Small chocolate smudges around the mouth, cheeks, and fingertips. A few subtle drops of melted chocolate on the floor. Keep everything cute, clean, charming, and controlled. Never messy.
ENVIRONMENT
Minimal background. Neutral surface. Soft natural daylight. Clean studio atmosphere. Strong visual contrast between miniature scale and real-world objects.
STYLE
Premium handcrafted collectible.
Ghibli-inspired charm.
Ultra-detailed sculpting.
Realistic miniature photography.
Physical artisan toy aesthetic.
Cute and heartwarming storytelling.
Luxury figurine product photography.
OUTPUT STYLE
Collectible designer toy.
Premium miniature sculpture.
Studio-quality photography.
Clean composition.
Photorealistic materials.
Highly detailed craftsmanship.
8K quality.
NEGATIVE PROMPT
Avoid excessive caricature. Avoid distorted facial features. Avoid flat printed faces. Avoid lifeless eyes. Avoid messy chocolate stains. Avoid hiding the chocolate. Avoid losing identity resemblance. Avoid deformed hands. Avoid duplicate characters. Avoid text. Avoid watermarks. Avoid cluttered backgrounds.
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 2,740 characters
Show the full prompt
Transform the uploaded image into a premium collectible 3D toy version while preserving the exact original scene, subject placement, pose, camera angle, framing, proportions, lighting direction, shadows, and overall composition.
FORMAT LOCK
Preserve the original aspect ratio, composition, crop, camera distance, perspective, subject scale, and scene structure. Do not redesign, reposition, remove, or add major elements. The final image must look like the same photograph converted into a toy world.
IDENTITY LOCK
Preserve the exact facial structure, facial proportions, hairstyle, hair color, body proportions, clothing design, expression, and recognizable identity traits. The subject must remain immediately recognizable.
TOY TRANSFORMATION
Convert all subjects into premium collectible designer toys. Ultra-smooth molded plastic surfaces. Clean glossy finish. Controlled specular highlights. Luxury figurine quality. High-end collectible aesthetic.
TOY CONSTRUCTION DETAILS
Visible articulation joints at shoulders, elbows, hips, and knees. Clean circular or segmented toy joints. Subtle seam lines along limbs and body parts. Optional minimal face segmentation lines inspired by premium action figures. Joints must feel integrated, realistic, and anatomically believable.
MATERIALS
Skin becomes smooth synthetic plastic with no pores or imperfections while preserving identity. Eyes become slightly glossy painted toy eyes. Clothing transforms into molded plastic, rubberized fabric, latex-like, or coated materials while preserving the exact original outfit design, folds, stitching, silhouette, and structure.
LIGHTING
Match the original image lighting exactly. Preserve the same lighting direction, intensity, shadow placement, and overall mood. Refine the lighting into premium CGI-quality rendering with soft highlight roll-off, clean reflections, balanced contrast, and realistic toy-material response. No dramatic HDR. No cinematic relighting.
RENDERING
Physically based rendering. Premium toy photography aesthetic. Clean reflections. Sharp molded details. Ultra-detailed surface quality. High-end product-photography realism.
OUTPUT STYLE
Luxury collectible figure. Designer toy aesthetic. Premium action-figure rendering. Photorealistic toy materials. High-end CGI craftsmanship. Studio-quality rendering. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid changing identity, changing pose, changing composition, changing camera angle, changing clothing design, missing joints, oversized joints, broken anatomy, realistic skin pores, matte clay textures, plush materials, cartoon rendering, low detail, low quality, text, watermark, duplicate subjects, extra limbs, distorted anatomy, or background redesign.
4 examples are not shown here because they show real public figures or show smoking. See all on the site ↗
Plush keychain character
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 1,980 characters
Show the full prompt
Transform the uploaded person into a cute premium plush keychain character while preserving the person’s recognizable facial identity, hairstyle, skin tone, and distinctive appearance in a simplified handcrafted form.
FORMAT LOCK
Vertical 3:4 composition. Close-up macro product photography. The miniature plush keychain is the main focal point, held gently between two realistic human fingers.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Translate the person’s recognizable facial features, hairstyle, expression, and overall appearance into a soft plush character design. Keep the likeness recognizable while adapting the features naturally to the simplified plush style.
PLUSH DESIGN
Create a small handcrafted keychain made from soft felt, fleece, and fuzzy fabric. Use cute rounded proportions, embroidered facial features, simple expressive eyes, stitched details, soft seams, tiny fabric fibers, and subtle handmade imperfections. Attach the plush securely to a shiny silver metal keyring.
POSE & HAND
The plush character is gently held between two realistic fingers, with natural finger positioning and believable scale. Keep the hand softly lit and slightly out of focus so the plush remains the primary subject.
BACKGROUND
Neutral warm beige studio background with generous negative space. Soft creamy bokeh and subtle depth separation.
LIGHTING & STYLE
Soft natural studio lighting. Warm gentle highlights. Shallow depth of field. Macro lens aesthetic. Highly detailed fabric and embroidery textures. Premium handcrafted product photography. Cute collectible aesthetic. Realistic material rendering. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid changing the recognizable appearance, plastic materials, hard CGI surfaces, realistic human face on the plush, distorted fingers, extra fingers, extra limbs, duplicate keyrings, blurry plush details, excessive decorations, cluttered background, text, logos, watermarks, or low-quality rendering.
1 example is not shown here because they show real public figures. See all on the site ↗
Bobblehead football figure
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 2,843 characters
Show the full prompt
Create a premium collectible bobblehead football figure inspired by the uploaded person while preserving only the general appearance, hairstyle silhouette, hair color, and visual character. Do not copy the real face exactly. Reinterpret the subject as a stylized toy collectible.
FORMAT LOCK
Vertical 4:5 composition. Centered figure. Full collectible visible. Physical product photography aesthetic. Front three-quarter camera angle.
BOBBLEHEAD DESIGN
Oversized vinyl head. Small football-player body. Authentic bobblehead proportions. Large spring-mounted wobble-head construction. Visible toy manufacturing details. Premium collectible figure aesthetic.
FACE DESIGN
Inspired by the uploaded person but redesigned as a stylized toy character. Simplified facial features. Cute collectible proportions. Avoid direct identity replication. Avoid realistic skin and realistic human likeness.
TOY MATERIALS
Molded vinyl construction. Visible mold seams. Minor paint imperfections. Small manufacturing flaws. Light scratches. Subtle wear marks. Slight paint bleed. Real physical toy appearance rather than CGI perfection.
FOOTBALL UNIFORM
National-team-inspired football kit based on {{Country}}.
Jersey uses {{Primary Color}} as the dominant color with {{Secondary Color}} details and {{Accent Color}} accents.
Shorts, socks, stripes, trim details, sleeve accents, collar details, and kit styling should reflect the national identity and color palette of {{Country}} while remaining authentic football-uniform design.
No copyrighted logos or official federation crests.
POSE
One foot positioned beside a football. Confident football-player stance. Static collectible-figure posture.
BASE
Premium football-pitch display base. Green grass texture with border colors inspired by {{Country}}.
ENVIRONMENT
Outdoor football field. Warm late-afternoon sunlight. Shallow depth of field. Soft blurred stadium or football-pitch background. Clean centered composition. Sharp ground shadow beneath the figure.
LIGHTING
Natural golden-hour lighting. Realistic product-photography shadows. Soft highlights on vinyl surfaces. Physical toy-photography realism.
STYLE
Premium bobblehead collectible. Physical product photography. Designer toy aesthetic. Sports memorabilia figure. High-end vinyl collectible. Real photographed object, not CGI.
OUTPUT STYLE
Ultra-realistic toy photography. Collectible football figurine. Professional product shot. Premium vinyl toy. Sharp focus on figure. Soft background blur. High-detail 8K quality.
NEGATIVE PROMPT
Avoid realistic human skin, realistic hair strands, exact identity replication, CGI rendering, anime style, cartoon illustration, action pose, oversized muscles, text, logos, watermarks, duplicate figures, low quality, floating objects, unrealistic materials, or digital-art appearance.
Make the aspect ratio 3:4
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 2,952 characters
Show the full prompt
Using the uploaded image as inspiration, create a premium collectible soccer-player toy sealed inside professional retail blister packaging.
FORMAT LOCK
Vertical 3:4 composition, full packaging visible, front-facing product photography, centered composition, premium toy-store shelf presentation.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve only the general appearance, hairstyle silhouette, hair color, and visual personality. Reinterpret the subject as an original collectible toy character. Do not replicate the real face or identity exactly.
COLLECTIBLE FIGURE
Premium vinyl toy, stylized football-player proportions, slightly oversized head, small athletic body, molded plastic hair with chunky sculpted grooves, simplified facial features, designer-toy aesthetic, original fictional character inspired by {{Country}}.
FOOTBALL UNIFORM
Create a completely original football kit inspired by {{Country}}. Use colors, patterns, and visual themes associated with the country's culture and flag colors. No team kits, federation logos, tournament branding, copyrighted jersey designs, recognizable athlete references, or sportswear-brand identities.
ACCESSORIES
Include several generic football accessories inside the blister packaging: football, training cones, shin guards, water bottle. Add two subtle humorous cultural references inspired by {{Country}}. Keep them playful, tasteful, and collectible-toy appropriate.
PACKAGING DESIGN
Premium retail blister packaging, glossy transparent blister shell, matte printed backing card, premium toy-brand presentation, collector-edition aesthetic, shelf-ready packaging design.
Large collectible title based on {{Country}}.
Country-inspired graphics, flag-color elements, collectible-design details, premium toy packaging aesthetics. No copyrighted logos or trademarks.
MATERIALS
Injection-molded plastic textures, premium vinyl surfaces, subtle manufacturing details, soft reflections, clean highlights, realistic packaging materials, physical toy appearance.
LIGHTING
Professional studio product photography, softbox reflections on blister plastic, clean shadows, commercial advertising quality, premium collectible-product lighting.
STYLE
Designer toy collectible, premium action figure, luxury toy packaging, retail display photography, high-end collectible merchandise, physical toy realism.
OUTPUT STYLE
Ultra-realistic product photography, premium blister-pack collectible, toy-store shelf aesthetic, professional packaging render, high-detail materials, commercial-quality advertising image, ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid real football clubs, federation logos, tournament branding, sportswear logos, athlete likenesses, copyrighted jersey designs, realistic human skin, realistic hair strands, CGI character rendering, text artifacts, watermark, duplicate accessories, damaged packaging, low quality, or cropped packaging.
Make the aspect ratio 3:4
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Travel posters
Full prompt, verbatim · 2,499 characters
Show the full prompt
Create a premium collectible travel poster in a cute stylized 3D cartoon aesthetic using the uploaded image as the character identity reference.
FORMAT LOCK
Vertical 3:4 poster composition, premium travel-poster design, centered Polaroid composition, social-media-ready artwork, high-resolution collectible illustration.
IDENTITY LOCK
Preserve the general facial identity, hairstyle, hair color, facial proportions, and recognizable appearance from the uploaded image while transforming the character into a premium stylized 3D cartoon figure.
POLAROID COMPOSITION
A charming {{Gender}} character inspired by {{Country}} is taking a cheerful smartphone selfie while sitting on the bottom edge of a large Polaroid frame.
One leg hangs naturally outside the Polaroid border creating a playful 3D pop-out effect, strong depth and dimensionality, natural interaction with the frame.
INSIDE THE POLAROID
A cozy travel scene inspired by {{Country}}.
Authentic architecture, food, landscapes, local lifestyle, nature, crafts, transportation, cultural details, and storytelling elements, warm inviting atmosphere, rich environmental details, modern fashion inspired by the colors and identity of {{Country}}, happy expression, large expressive eyes, detailed stylized hair, appealing collectible-character design, selfie-camera perspective energy.
OUTSIDE THE POLAROID
Soft premium gradient background inspired by {{Country}}, elegant monochrome travel doodles, landmarks, local foods, animals, plants, transportation icons, maps, cultural symbols, decorative travel motifs, clean layout with generous negative space, no clutter.
TYPOGRAPHY
Large handwritten country name at the top, small national flag beside the title, short country slogan beneath, inspirational handwritten quote placed naturally on the Polaroid border.
STYLE
Premium stylized 3D illustration, collectible travel-souvenir poster, cute character design, Pixar-inspired charm, soft cinematic lighting, detailed textures, depth of field, vibrant but tasteful colors, premium editorial composition.
OUTPUT STYLE
High-end 3D travel poster, collectible souvenir artwork, premium social-media illustration, ultra-cute character design, rich storytelling environment, professional composition, ultra-detailed 4K quality.
NEGATIVE PROMPT
Avoid stereotypes, cluttered compositions, exaggerated caricatures, low-quality rendering, text errors, watermark, flat lighting, messy layouts, distorted anatomy, duplicate objects, generic travel imagery.
3 examples are not shown here because they show real public figures. See all on the site ↗
You and your big-head 3D character twin
Figure potential: high (big-head stylized character suits a printed figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 3,750 characters
Show the full prompt
Use the uploaded reference image as the immutable identity source for both subjects.
IDENTITY LOCK (NON-NEGOTIABLE)
Preserve the exact facial structure, facial proportions, skin tone, eye shape, nose shape, lips, hairstyle, hair color, age appearance, body proportions, and all recognizable identity traits from the reference image.
The real person and the stylized character must clearly represent the same individual.
Do not generate a different face, different hairstyle, different ethnicity, different age, or different body type.
Do not beautify, redesign, or reinterpret the person's identity.
COMPOSITION & FRAMING
Create a premium studio portrait in a vertical 4:5 composition.
Full-body shot.
The real person stands on the right side of the frame.
A stylized 3D character version of the same person stands on the left side.
The 3D character should be approximately 50–60% of the real person's height, reaching roughly waist level.
The real subject casually rests one elbow on top of the character's oversized head.
The interaction should feel natural, relaxed, and believable.
Maintain balanced spacing and a clean centered composition.
REAL SUBJECT (PHOTOREALISTIC)
Create an ultra-realistic studio photograph with professional commercial quality.
Visible skin texture, natural pores, realistic hair strands, authentic fabric details, and physically accurate lighting.
Expression should feel warm, approachable, confident, and relaxed.
Avoid exaggerated smiles, artificial beauty filters, plastic skin, or unrealistic retouching.
POSE DIRECTION
The pose must feel naturally stylish and effortless.
Weight slightly shifted onto one leg.
Subtle asymmetry in posture.
Relaxed shoulders.
Natural body language.
The arm resting on the character should feel casual rather than posed.
Avoid stiff standing positions.
CLOTHING CONTINUITY
Recreate the outfit from the reference image exactly.
Preserve garment colors, fit, silhouette, fabric characteristics, stitching, folds, seams, cuffs, hems, logos, accessories, footwear, and overall styling.
Do not redesign or substitute clothing items.
3D CHARACTER VERSION
Create a premium animated 3D character inspired by high-end cinematic animation aesthetics.
The character must remain clearly recognizable as the same individual.
Large expressive head.
Smaller stylized body proportions.
Enhanced facial expressions while maintaining identity accuracy.
Slightly larger eyes.
Softer facial contours.
Friendly animated smile.
Refined high-quality character modeling.
Avoid toy-like, cheap, plastic, low-detail, or rigid figurine appearances.
CHARACTER PERSONALITY & MOTION
The character should feel alive and expressive.
Subtle body movement.
Natural stance.
Playful personality.
Animated facial expression.
Confident and charming presence.
The pose should feel like a frame captured from an animated film rather than a static collectible figure.
MATERIALS & DETAILING
High-quality physically based rendering.
Premium fabric simulation.
Soft realistic clothing folds.
Accurate surface reflections.
Natural skin shading.
Detailed hair rendering.
Professional character texturing.
LIGHTING & STUDIO ENVIRONMENT
Soft diffused studio lighting from the upper-left.
Clean highlight transitions.
Natural soft shadows.
Premium commercial photography lighting setup.
Background should be a seamless light-gray studio gradient, slightly brighter toward the center.
Both subjects must cast realistic grounded shadows.
OUTPUT STYLE
Luxury commercial studio portrait.
Ultra-clean composition.
Photorealistic human subject paired with a premium stylized 3D character.
Extremely high detail.
Cinematic quality.
Professional advertising campaign aesthetic.
Sharp focus.
Natural depth.
Magazine-level production quality.
7 examples are not shown here because they show real public figures. See all on the site ↗
Stylized 3D caricature character
Figure potential: high (big-head stylized character suits a printed figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 3,178 characters
Show the full prompt
Create a premium stylized 3D caricature character based on the uploaded reference image.
IDENTITY LOCK — NON-NEGOTIABLE
Use the uploaded image as the exact identity source.
Preserve the subject's facial identity, facial structure, skin tone, eye shape, nose shape, lips, hairstyle, hair color, age appearance, and all recognizable features.
The character must remain immediately recognizable as the same person.
Do not generate a different face, hairstyle, ethnicity, age, or facial proportions.
Do not beautify, redesign, or reinterpret the person's identity.
CHARACTER DESIGN
Transform the subject into a premium stylized 3D caricature character.
Oversized head.
Smaller full upper body.
Slightly elongated neck.
Large expressive eyes with soft eyelids.
Fuller stylized lips.
Subtle facial asymmetry for realism and personality.
Human-inspired but intentionally exaggerated.
Premium designer-toy character style.
The character must have a visible torso, shoulders, arms, and clothing.
Do not create a bust statue.
Do not cut the body into a sculpture base.
Do not add a pedestal, stand, platform, circular base, display base, museum base, or collectible base.
BODY & COMPOSITION
Show the character from head to upper body or half-body.
The character should appear as a standing stylized person, not a statue.
Keep the body naturally connected below the neck.
Include shoulders, chest, arms, and outfit.
No floating head.
No bust sculpture.
No mannequin base.
No black circular base.
No object under the body.
Clean studio framing similar to a premium 3D character render.
SURFACE & MATERIALS
Ultra-clean skin surface.
Smooth premium material finish.
Refined facial detailing.
Pristine collectible-quality craftsmanship.
Polished modern surfaces.
High-end designer character appearance.
Avoid skin noise, freckles, dirt, smudges, texture artifacts, grain, scratches, or imperfections.
FACIAL EXPRESSION
Relaxed confident expression.
Subtle personality.
Stylish modern attitude.
Soft natural emotion.
Avoid exaggerated cartoon expressions.
Avoid extreme smiling.
LIGHTING
Soft studio lighting.
Balanced neutral illumination.
Clean shadow transitions.
Gentle natural shadows.
Even exposure.
BACKGROUND
Simple neutral gray backdrop.
Warm off-white studio background.
Minimal environment.
No environmental distractions.
OUTPUT STYLE
Premium stylized 3D caricature character.
Full upper-body character render.
Designer-toy aesthetic without a base.
Ultra-detailed rendering.
Hyper-realistic materials.
Luxury studio character portrait.
Pristine finish.
Modern character sculpture feel, but not a bust statue.
Sharp focus.
Ultra-high resolution.
NEGATIVE PROMPT
Avoid pedestal, stand, platform, circular base, black base, display base, museum base, bust statue, sculpture bust, cut-off torso, floating head, mannequin base, trophy base, cartoon style, Pixar style, anime style, toy-store appearance, low-detail sculpting, distorted anatomy, unrealistic proportions, blurry details, damaged textures, dirty materials, exaggerated expressions, text, logos, branding, watermark, low resolution, CGI artifacts, overexposed lighting, cinematic backgrounds, and unrealistic skin rendering.
Figure potential: high (big-head stylized character suits a printed figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 3,362 characters
Show the full prompt
Create a premium high-quality 3D caricature companion of the person in the uploaded image while keeping the original person fully present, photorealistic, and completely unchanged.
FORMAT LOCK
Preserve the original aspect ratio, framing, composition, camera angle, crop, perspective, lighting, background, and scene structure.
REAL PERSON LOCK — CRITICAL
The original person must remain untouched.
Preserve exact facial features, face shape, hairstyle, hair color, eyebrows, eyes, nose, lips, jawline, skin tone, expression, clothing, accessories, pose, body proportions, and overall identity.
Do not stylize, cartoonize, replace, modify, redesign, beautify, or alter the real person in any way.
CARICATURE COMPANION SYSTEM
Create a premium stylized 3D caricature version of the exact same person standing naturally beside their real-life counterpart.
The caricature must clearly resemble the same person.
Preserve recognizable identity, hairstyle, outfit, accessories, fashion styling, attitude, personality, and overall vibe.
PROPORTIONS
{{Height Ratio}} height relative to the real person.
The caricature should reach approximately shoulder-to-ear height of the real person.
Never tiny, Never mascot-sized, Never toy-sized, Never child-sized.
Large expressive eyes, Slightly oversized head, Smaller body proportions, Elegant stylized anatomy, Premium animated-character proportions.
INTERACTION
{{Interaction}}
The caricature should stand naturally beside the real person.
Subtle physical interaction only.
Examples: lightly touching the person's shoulder, gently leaning toward the person, resting a hand on the person's upper arm, standing close with natural body language, looking toward the person, walking side-by-side, sharing the same direction of attention.
Avoid hand holding, Avoid hugging, Avoid exaggerated interaction.
The relationship should feel warm, natural, candid, and believable.
The caricature should feel like a living animated version of the same person existing naturally within the real world.
MATERIALS
Premium cinematic 3D rendering, Smooth stylized skin, Luxury animation quality, Detailed hair sculpting, High-end character materials, Not plastic, Not toy-like, Not vinyl.
STYLE
{{Caricature Style}} stylized realism.
Pixar-inspired quality, Premium animated-film aesthetic, Luxury character design, Sophisticated and fashionable appearance, Strong likeness retention.
CARICATURE MOOD
{{Caricature Mood}}
Warm personality, Natural expression, Friendly presence, Subtle emotional connection with the real person.
LIGHTING
Match the original scene lighting exactly, Consistent shadows, Consistent reflections, Natural environmental integration.
The caricature must appear physically present in the same environment as the real person.
OUTPUT STYLE
Photorealistic human with premium animated companion, cinematic realism, luxury editorial photography, high-end 3D character design, seamless real-world integration, ultra-detailed rendering, 8K quality.
NEGATIVE PROMPT
Tiny companion, mascot proportions, toy figure, Funko Pop, chibi style, baby proportions, hand holding, hugging, disconnected character, floating character, separate scene, plastic materials, exaggerated cartoon style, unrealistic interaction, altered real person, duplicated real person, low quality, watermark, text, distorted anatomy.
Make the aspect ratio 3:4.
Figure potential: high (big-head stylized character suits a printed figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 1,778 characters
Show the full prompt
Transform the uploaded person into a highly stylized premium 3D caricature while preserving their recognizable identity, facial structure, hairstyle, skin tone, expression, outfit, colors, and overall appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body character. Centered composition. Minimal clean studio background. Keep the entire character visible.
IDENTITY LOCK
Preserve the exact facial features, eye shape, nose, lips, jawline, hairstyle, skin tone, body proportions, clothing, shoes, and recognizable likeness. Do not redesign the person or create a different identity.
CHARACTER STYLE
Create an oversized expressive head with playful caricature proportions, slightly smaller body, expressive eyes, soft facial exaggeration, rounded forms, and a charming confident personality. Keep the face recognizable while enhancing its most distinctive features naturally.
3D RENDERING
Smooth premium 3D materials. Soft realistic skin texture. Clean polished surfaces. Subtle fabric detail. Soft ambient studio lighting. Gentle contact shadows. Natural highlights. High-quality cinematic 3D character rendering.
BACKGROUND
Minimal seamless studio background with generous negative space. Soft neutral tones. Keep the environment simple so the character remains the clear focal point.
OUTPUT STYLE
Premium 3D character design. Stylized caricature. Modern animated-film aesthetic. Polished materials. Expressive face. Clean composition. Professional studio render. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid distorted identity, unrecognizable face, excessive facial exaggeration, plastic skin, stiff pose, low detail, blurry rendering, extra limbs, extra fingers, bad anatomy, cluttered background, text, logos, watermarks, or low-quality CGI.
UPLOAD A REFERENCE IMAGE BEFORE GENERATING.
MAIN PROMPT
Using the uploaded photo as the primary reference, transform ONLY the main subject into a charming oversized-head hand-drawn cartoon caricature. The head must be dramatically exaggerated, approximately 45–60% of the character’s visual height, with a compact, much smaller body. Preserve the subject’s recognizable identity and defining characteristics while keeping the original photograph and environment fully recognizable.
FORMAT LOCK
Vertical 3:4. Preserve the original framing, camera angle, perspective, depth, subject placement, pose, orientation, and composition. Keep the photographic environment unchanged.
STYLE TRANSFER
Create a naïve children’s-book caricature with a deliberately huge, subject-specific head, enlarged forehead and cheeks, tiny simplified body, short limbs, small hands and feet, and a cute humorous proportion. Adapt the head shape, silhouette, hairstyle, hairline, facial hair, ears, beak, hat, helmet, or other defining features specifically to the original subject. Never use a generic circular head.
DETAIL & COLOR
Preserve facial structure, expression, skin tone, hairstyle, facial hair, clothing, accessories, patterns, logos, shoes, jewelry, and distinctive details. Use simple dot/oval eyes, tiny simplified nose, curved mouth, subtle cheeks when appropriate, soft natural colors, visible colored-pencil strokes, rough hatching, cross-hatching, uneven fills, overlapping strokes, imperfect edges, and subtle paper texture.
COMPOSITION LOCK
Transform ONLY the main subject. Keep the entire background photorealistic, including architecture, landscape, furniture, vehicles, scenery, lighting, depth, perspective, and atmosphere. Integrate the illustrated character naturally into the original scene with matching position, perspective, lighting direction, and subtle grounding shadows.
STYLE
Loose hand-drawn ink and colored-pencil caricature, charming, whimsical, naïve, expressive, handmade, slightly imperfect. Thin-to-medium organic black outlines, rough pencil texture, spontaneous sketch marks, simplified clothing, exaggerated oversized head, tiny body. The contrast between the realistic photograph and illustrated subject should be intentional and visually striking.
NEGATIVE PROMPT
Normal head proportions, small head, generic head shape, generic cartoon character, identical head shape across subjects, photorealistic subject, cartoonized background, fully illustrated scene, anime, vector art, 3D CGI, Pixar, Disney, glossy digital rendering, plastic texture, smooth gradients, polished graphics, altered identity, changed hairstyle, changed clothing, changed pose, distorted facial features, excessive facial detail, oversized body, realistic anatomy, unnecessary background changes.
ASPECT RATIO: 3:4
Figure potential: high (big-head stylized character suits a printed figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,678 characters
Show the full prompt
Transform the uploaded photo into an intentionally very ugly, minimalist Indonesian caricature while keeping the person’s recognizable identity as the highest priority. Preserve the original composition and visual information while simplifying everything into a crude, amateur hand-drawn caricature.
FORMAT LOCK
Vertical 3:4 composition, 1080×1350. Preserve the exact original framing, crop, camera angle, perspective, pose, gesture, expression, clothing, accessories, and subject placement. Do not zoom, rotate, reposition, expand, or invent missing body parts.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the person’s recognizable facial structure, hairstyle or hijab, hairline, glasses, facial hair, expression, skin tone, and distinctive features. The person must remain instantly recognizable before exaggeration.
UGLY CARICATURE STYLE
Make the drawing deliberately crude, awkward, uneven, and charmingly amateur, as if created by someone with very limited drawing skills. Use an oversized head occupying roughly 45–50% of the visible body, wide cheeks, a square jaw narrowing into a long pointed chin, tiny sleepy oval eyes with dot pupils, thick uneven eyebrows, a tiny one-stroke doodle nose, oversized uneven lips, crude teeth if visible, flat silhouette hair, and thick slightly crooked black glasses.
LINE & COLOR
Use shaky uneven black outlines with flat opaque colors only. Flat peach-beige skin. Simplified flat clothing colors with only a few crude wrinkle lines. Simplify all visible objects using the same intentionally ugly drawing language. No gradients, shadows, highlights, textures, realistic rendering, or polished vector precision.
BACKGROUND
Replace the entire background with one solid pastel color inspired by the dominant mood of the original image. Completely flat and empty. No scenery, furniture, decorations, textures, patterns, or gradients.
STYLE PRIORITY
Identity first, composition second, ugly simplicity third. Always choose the simplest crude version that still preserves the person’s recognizable identity. The result should feel intentionally bad, funny, handmade, minimal, and charming rather than professionally illustrated.
NEGATIVE PROMPT
Avoid beautiful illustration, anime, Pixar, Disney, 3D, semi-realistic rendering, digital painting, realistic anatomy, realistic hair strands, realistic nose, cinematic lighting, beauty enhancement, polished vector art, smooth professional linework, gradients, shadows, textures, extra limbs, extra fingers, invented body parts, altered pose, changed crop, expanded composition, zooming, rotation, repositioning, background details, text, logos, or watermarks.
Figure potential: high (already a buildable toy form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 4,494 characters
Show the full prompt
Transform every visible person in the uploaded image into realistic LEGO-style minifigure toys while preserving each person’s recognizable identity, hairstyle, outfit, colors, accessories, relative height, pose, and relationship to each other.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 3:4 aspect ratio.
Full-body product composition.
All figures must be fully visible from head to toe.
Do not generate square, landscape, widescreen, close-up, cropped, or half-body compositions.
PERSON COUNT LOCK — NON-NEGOTIABLE
Detect all visible people in the uploaded image.
If there is 1 person, create 1 LEGO-style minifigure.
If there are 2 people, create 2 LEGO-style minifigures.
If there are 3 or more people, create the same number of LEGO-style minifigures.
Do not remove, merge, ignore, duplicate, or replace any person.
Every person visible in the reference image must appear as their own separate LEGO-style minifigure.
REFERENCE LOCK — NON-NEGOTIABLE
Use the uploaded image as the only reference for all subjects.
Preserve each person’s recognizable facial identity as much as possible within the simplified LEGO minifigure style.
Maintain each person’s hairstyle and hair color, adapted into molded LEGO-style plastic hairpieces.
Preserve each person’s facial expression, clothing colors, clothing design, visible patterns, accessories, and styling.
The final figures must clearly represent the same people translated into LEGO-style toy form.
Do not create generic minifigures.
Do not redesign outfits or invent new accessories.
CLOTHING DETAIL LOCK
Recreate each person’s outfit as accurately as possible in LEGO minifigure form.
Preserve the exact clothing color palette, layering, garment type, collar shape, neckline, sleeves, jacket shape, pants or skirt style, shoes, visible patterns, seams, buttons, zippers, belts, logos, jewelry, bags, glasses, hats, and accessories.
Translate real clothing into printed LEGO torso and leg graphics with clean toy-style linework.
Outerwear should appear as molded or printed LEGO-style clothing details.
Shoes should be simplified into LEGO foot/leg prints while keeping the original color and style.
Do not replace outfits with generic LEGO clothing.
Do not simplify clothing so much that the original styling is lost.
GROUP COMPOSITION
Keep the same number of people from the uploaded image.
Preserve the relationship between subjects.
Preserve relative height differences.
Preserve basic pose direction and body language.
If the reference shows a couple, create a matching LEGO-style couple.
If the reference shows a group, create the full group as separate minifigures.
Arrange the figures naturally together in a clean product-style composition.
LEGO MINIFIGURE STRUCTURE
Use authentic LEGO-style minifigure proportions for every figure.
Cylindrical heads.
Simplified printed facial features.
Block-shaped torsos.
Printed clothing details on each torso.
Standard minifigure arms.
Curved minifigure hands.
Short minifigure legs.
Custom molded plastic hairpieces matching each subject’s hairstyle.
Classic yellow heads or realistic skin-tone heads depending on the uploaded subjects and chosen visual direction.
MATERIAL & TEXTURE
Render all figures as realistic molded plastic.
Subtle glossy finish.
Fine molded seams.
Clean toy-surface reflections.
Realistic plastic highlights.
Printed face, torso, and leg details.
Hairpieces should look like molded plastic, not real hair.
Clothing should appear as printed LEGO-style graphics and molded toy details, not real fabric.
COMPOSITION
Full-body minifigures visible from head to toe.
All figures centered in the frame.
Entire group fully visible.
No cropping.
No close-up portrait framing.
No props.
No scenery.
No extra elements.
BACKGROUND
Pure clean white background.
Studio product photography setup.
Minimal and premium.
No environment, no table clutter, no unrelated objects.
LIGHTING
Soft diffused studio lighting.
Subtle natural shadows beneath the figures.
Crisp detail visibility.
Balanced reflections across the plastic surfaces.
Professional toy photography quality.
OUTPUT STYLE
Photorealistic LEGO-style minifigure toy render.
High-quality 3D product visualization.
Sharp focus.
Clean white-background presentation.
Premium collectible toy aesthetic.
Highly detailed plastic texture.
Professional studio lighting.
The final image should look like realistic custom LEGO-style minifigures of every visible person from the uploaded image, photographed as a premium toy product.
Figure potential: high (already a buildable toy form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 5,961 characters
Show the full prompt
Transform every visible person in the uploaded image into realistic LEGO-style minifigure toys while preserving each person’s recognizable identity, hairstyle, outfit, colors, accessories, relative height, pose, and relationship to each other.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 3:4 aspect ratio. Full-body product composition. All figures must be fully visible from head to toe. Do not generate square, landscape, widescreen, close-up, cropped, or half-body compositions.
PERSON COUNT LOCK — NON-NEGOTIABLE
Detect all visible people in the uploaded image. If there is 1 person, create 1 LEGO-style minifigure, if there are 2 people, create 2 LEGO-style minifigures, if there are 3 or more people, create the same number of LEGO-style minifigures. Do not remove, merge, ignore, duplicate, or replace any person. Every person visible in the reference image must appear as their own separate LEGO-style minifigure.
REFERENCE LOCK — NON-NEGOTIABLE
Use the uploaded image as the only reference for all subjects. Preserve each person’s recognizable facial identity as much as possible within the simplified LEGO minifigure style. Maintain each person’s hairstyle and hair color, adapted into molded LEGO-style plastic hairpieces. Preserve each person’s facial expression, clothing colors, clothing design, visible patterns, accessories, and styling. The final figures must clearly represent the same people translated into LEGO-style toy form. Do not create generic minifigures. Do not redesign outfits or invent new accessories.
CLOTHING DETAIL LOCK
Recreate each person’s outfit as accurately as possible in LEGO minifigure form. Preserve the exact clothing color palette, layering, garment type, collar shape, neckline, sleeves, jacket shape, pants or skirt style, shoes, visible patterns, seams, buttons, zippers, belts, logos, jewelry, bags, glasses, hats, and accessories. Translate real clothing into printed LEGO torso and leg graphics with clean toy-style linework. Outerwear should appear as molded or printed LEGO-style clothing details. Shoes should be simplified into LEGO foot/leg prints while keeping the original color and style. Do not replace outfits with generic LEGO clothing. Do not simplify clothing so much that the original styling is lost.
GROUP COMPOSITION
Keep the same number of people from the uploaded image. Preserve the relationship between subjects. Preserve relative height differences. Preserve basic pose direction and body language. If the reference shows a couple, create a matching LEGO-style couple. If the reference shows a group, create the full group as separate minifigures. Arrange the figures naturally together in a clean premium product-style composition.
LEGO MINIFIGURE STRUCTURE
Use authentic LEGO-style minifigure proportions for every figure. Cylindrical heads, simplified printed facial features, block-shaped torsos, printed clothing details on each torso, standard minifigure arms, curved minifigure hands, short minifigure legs, custom molded plastic hairpieces matching each subject’s hairstyle. Classic yellow heads or realistic skin-tone heads depending on the uploaded subjects and chosen visual direction.
MATERIAL & TEXTURE
Render all figures as realistic molded plastic. Subtle glossy finish. Fine molded seams. Clean toy-surface reflections. Realistic plastic highlights. Printed face, torso, and leg details. Hairpieces should look like molded plastic, not real hair. Clothing should appear as printed LEGO-style graphics and molded toy details, not real fabric.
COMPOSITION
Full-body minifigures visible from head to toe. All figures centered in the frame. Entire group fully visible. No cropping. No close-up portrait framing. No props. No scenery. No extra elements.
BACKGROUND
Use a bold, vibrant LEGO-inspired red studio background. Create a seamless saturated red backdrop inspired by the iconic red associated with LEGO branding and packaging. The red should be bright, energetic, playful, clean, and premium. Use a very subtle tonal gradient: slightly deeper red toward the outer edges with a gently brighter red behind the figures. No LEGO bricks. No LEGO logo. No text. No patterns. No scenery. No props. No unrelated objects. Keep the entire background minimal and uncluttered. Add subtle soft contact shadows beneath the figures.
LIGHTING
Soft diffused professional studio lighting. Crisp detail visibility. Balanced reflections across the plastic surfaces. Soft highlights on the molded plastic. Natural contact shadows beneath the figures. Use lighting that separates the colorful minifigures clearly from the red background without creating harsh glare.
OUTPUT STYLE
Photorealistic LEGO-style minifigure toy render. High-quality 3D product visualization. Sharp focus. Premium collectible toy photography. Vibrant LEGO-inspired red studio presentation. Highly detailed molded plastic texture. Professional commercial product photography. The final image should look like realistic custom LEGO-style minifigures of every visible person from the uploaded image, photographed as a premium collectible toy product against a bold LEGO-inspired red studio background.
NEGATIVE PROMPT
White background, gray background, black background, blue background, dull red, dark muddy red, busy background, LEGO bricks in background, LEGO logo, text, scenery, props, tables, furniture, extra objects, extra people, duplicate figures, missing people, merged people, generic characters, altered outfits, invented accessories, incorrect clothing colors, incorrect hairstyles, distorted faces, face drift, wrong person count, cropped bodies, half-body figures, missing feet, missing legs, missing heads, oversized heads, realistic human bodies, realistic fabric, real hair, cartoon people, anime style, low-detail plastic, blurry image, low resolution, distorted anatomy, floating figures, excessive shadows, harsh reflections, clutter, square composition, landscape composition, widescreen composition.
Figure potential: high (already a buildable toy form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Full prompt, verbatim · 2,309 characters
Show the full prompt
Transform the uploaded person into a premium ultra-realistic 3D LEGO-inspired room scene while preserving the person’s recognizable facial identity, facial structure, hairstyle, skin tone, body proportions, and overall appearance.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. The person sits naturally on a LEGO couch. Entire scene visible with balanced framing and strong miniature-world depth.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the person’s recognizable facial features, hairstyle, skin tone, expression, and natural body proportions. Do not redesign or replace the person’s identity.
LEGO WORLD
Create an entire room completely constructed from LEGO bricks. Every visible object and surface must be made from LEGO elements, including walls, floor, couch, table, lamp, shelves, plants, picture frames, mug, books, decorations, and small accessories. Use realistic LEGO brick connections, studs, seams, edges, and authentic plastic material.
CHARACTER
Dress the person in a fully LEGO-textured outfit with a jacket and trousers constructed from colorful LEGO pieces, plus LEGO-style sneakers. Integrate small LEGO pieces naturally into the hairstyle while keeping the person’s recognizable hairstyle and facial identity. Use a modern, stylish appearance with playful LEGO-inspired details.
ENVIRONMENT
Make the room vibrant, playful, and imaginative with colorful LEGO details throughout. Add layered furniture, shelves, plants, books, framed artwork, small decorations, and realistic LEGO construction details while keeping the composition organized and visually readable.
LIGHTING & STYLE
Soft professional studio lighting. Realistic shadows. Subtle reflections on LEGO plastic. Detailed brick textures. Authentic LEGO-like proportions and construction. Premium cinematic 3D rendering. Photorealistic material quality. High dynamic range. Sharp focus. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid changing the person’s identity, distorted facial features, plastic-looking skin, generic plastic blocks, missing LEGO studs, floating bricks, disconnected objects, extra limbs, distorted hands, duplicate objects, messy composition, blurry details, low quality, flat illustration, cartoon-only rendering, text, logos, watermarks, or cropped subject.
Figure potential: high (already a buildable toy form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 3,504 characters
Show the full prompt
Transform the uploaded real-life photo into a high-quality cinematic Minecraft-inspired world while preserving the exact original composition, framing, camera angle, crop, zoom level, subject placement, pose, perspective, and overall scene layout.
FORMAT LOCK — NON-NEGOTIABLE
Preserve the original image aspect ratio.
Preserve the original image dimensions.
Do not change the framing, crop, zoom level, camera distance, subject scale, or subject placement.
Do not generate a new composition.
The transformed image must match the original photo structure exactly.
SUBJECT LOCK — NON-NEGOTIABLE
Preserve the main human subject’s pose, body proportions, facial direction, hairstyle silhouette, clothing structure, body positioning, and perspective from the uploaded image.
The subject must remain clearly recognizable as the same person from the original photo.
Do not change the person’s identity, outfit, pose, body shape, hairstyle, or placement.
STYLE DIRECTION
Convert the human subject into a cinematic Minecraft-inspired character.
Use Minecraft-style block geometry while keeping the original person recognizable.
Clothing should become voxel-textured while preserving the original outfit design, colors, silhouette, and layering.
Face and body should feel like a high-end Minecraft RTX character, not a low-quality pixel filter.
Use realistic shader rendering, cinematic Minecraft lighting, and detailed voxel surface textures.
ENVIRONMENT TRANSFORMATION
Recreate the entire background as a Minecraft-inspired voxel world.
Terrain becomes cubic Minecraft blocks.
Trees become blocky voxel trees.
Grass, foliage, rocks, roads, paths, clouds, buildings, mountains, and landscapes become Minecraft-style block structures.
Water becomes pixelated Minecraft water with realistic shader reflections if present.
Preserve the original environmental depth, horizon placement, perspective, and spatial layout.
OBJECT & PET RULE
Any non-human object or animal in direct physical contact with the subject must also be converted into a Minecraft-style object or mob.
Dogs should become Minecraft-style wolves.
Nearby pets, bags, props, vehicles, benches, fences, or accessories should match the Minecraft voxel environment while preserving their original placement and function.
LIGHTING & ATMOSPHERE
Use realistic Minecraft RTX lighting.
Add cinematic fog and atmospheric depth when suitable.
Use voxel shadows and block-based light behavior.
Lighting should match the direction and mood of the original photo while becoming Minecraft shader-style.
If the original photo has haze, forest fog, sunset light, city glow, or mountain atmosphere, recreate it with cinematic Minecraft-style volumetric effects.
CAMERA & REALISM
Maintain the exact camera angle and framing from the reference photo.
Keep the same lens feeling, subject distance, and perspective.
The result should feel like the original photograph was directly transformed into a realistic Minecraft universe.
Avoid changing the scene layout or inventing a new background.
OUTPUT STYLE
High-quality cinematic Minecraft screenshot.
Ultra-detailed voxel world rendering.
Realistic Minecraft RTX gameplay aesthetic.
Block-textured terrain.
Authentic Minecraft visual identity.
Cinematic shader effects.
Stylized but believable game realism.
Immersive, atmospheric, and detailed.
The final image should feel like a real high-end Minecraft gameplay screenshot created from the uploaded photo, while preserving the original image structure exactly.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 3,333 characters
Show the full prompt
Transform the uploaded image into a highly detailed absurdist claymation world while preserving the exact original composition, object placement, perspective, camera angle, framing, proportions, and scene structure.
FORMAT LOCK
Preserve the original aspect ratio, composition, crop, camera distance, perspective, subject scale, and environmental layout. Do not move, remove, replace, redesign, crop, or reposition any major element. The final image must match the original image structure exactly.
CLAYMATION TRANSFORMATION
Convert the entire scene into a handcrafted stop-motion clay animation universe. Every person, object, animal, building, prop, and environmental element should become sculpted from physical plasticine clay. Nothing should remain realistic. Everything should feel physically handmade.
CHARACTER DESIGN
Extreme Aardman-inspired caricature style with an absurdist twist. Wildly exaggerated facial features, bulging ping-pong-ball eyes pointing slightly in different directions, massive toothy grins, oversized chiclet teeth, visible pink gums, exaggerated noses, expressive eyebrows, oversized cheeks, awkward facial proportions. Every character should appear hilariously startled, overexcited, confused, awkward, or absurdly enthusiastic.
BODY PROPORTIONS
Long noodle-like rubbery limbs, oversized heads, chunky hands, tiny feet, uneven proportions, soft rounded forms, exaggerated silhouettes, subtle clay deformation throughout the body.
CLAY TEXTURE SYSTEM
Heavy matte plasticine material. Deep thumbprints, visible finger impressions, accidental fingernail scratches, rough seams, dents, clay smears, tiny dust particles, lint fibers, hand-sculpted imperfections, uneven surfaces. Nothing should feel clean, polished, or machine-made.
GEOMETRY RULES
No perfectly straight lines. No perfect circles. No perfect symmetry. Introduce subtle wobble, handmade irregularities, lopsided construction, wonky geometry, soft deformation, and sculpted imperfections throughout the entire scene.
ENVIRONMENT TRANSFORMATION
Transform the entire environment into clay. Buildings become handmade clay structures, furniture becomes miniature clay models, roads become sculpted clay surfaces, plants become clay sculptures, props become handcrafted plasticine objects. Preserve the original scene layout exactly while transforming every surface into tactile clay.
LIGHTING
Professional stop-motion studio lighting, soft cinematic shadows, subtle clay highlights, miniature photography realism, warm handcrafted atmosphere, realistic physical-material response.
STYLE
Premium stop-motion animation, handcrafted clay sculpture aesthetic, absurdist character design, tactile plasticine materials, museum-quality miniature craftsmanship, whimsical comedy, classic clay-animation energy.
OUTPUT STYLE
Big-budget stop-motion movie frame, absurdist claymation artwork, handcrafted miniature universe, premium plasticine animation, ultra-detailed tactile surfaces, professional studio-quality rendering, ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid photorealism, realistic skin, realistic textures, perfect geometry, straight edges, symmetry, clean CGI surfaces, polished materials, realistic anatomy, Pixar style, anime style, low detail, low quality, text, watermarks, duplicated objects, missing limbs, or altered composition.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 3,095 characters
Show the full prompt
Transform the uploaded image into a highly detailed absurdist claymation world while preserving the exact original composition, spatial layout, object positioning, perspective, framing, and scene structure.
FORMAT LOCK — NON-NEGOTIABLE
Preserve the original aspect ratio.
Preserve the original composition.
Preserve the original camera angle.
Preserve the original object placement.
Preserve the original subject placement.
Do not crop, zoom, move, remove, or redesign any major element.
The final image must match the original image structure exactly.
CLAYMATION TRANSFORMATION
Convert the entire scene into a handcrafted stop-motion clay animation universe.
Every person, object, animal, prop, architectural element, and environmental detail should become sculpted from physical plasticine clay.
Nothing should remain realistic.
Everything should feel hand-built and physically sculpted.
CHARACTER DESIGN
Extreme Aardman-inspired caricature style.
Wildly exaggerated facial features.
Bulging asymmetrical eyes.
Slightly cross-eyed expressions.
Massive toothy grins.
Oversized cartoon teeth.
Visible pink gums.
Large noses.
Expressive eyebrows.
Exaggerated cheeks.
Every character should appear hilariously surprised, overexcited, awkward, or absurdly enthusiastic.
BODY PROPORTIONS
Long noodle-like limbs.
Rubbery arms and legs.
Slightly oversized heads.
Chunky hands.
Soft rounded forms.
Comically uneven proportions.
Natural clay deformation throughout the body.
CLAY TEXTURE SYSTEM
Heavy matte plasticine material.
Visible thumbprints.
Finger impressions.
Fingernail scratches.
Hand-sculpted imperfections.
Soft clay seams.
Small dents.
Dust particles.
Tiny lint fibers.
Uneven clay surfaces.
Nothing should feel machine-made or perfectly clean.
GEOMETRY RULES
No perfectly straight lines.
No perfect circles.
No perfect symmetry.
Introduce subtle wobble and handmade irregularities throughout the scene.
All objects should feel slightly wonky and imperfect.
The world should look physically sculpted by hand.
ENVIRONMENT
Transform the entire environment into clay.
Buildings become clay structures.
Furniture becomes clay models.
Nature becomes clay sculptures.
Roads become sculpted clay surfaces.
Props become handmade miniature clay objects.
Maintain the original scene layout while transforming every surface into tactile clay.
LIGHTING
Professional stop-motion studio lighting.
Soft cinematic shadows.
Warm handcrafted atmosphere.
Subtle clay surface highlights.
Realistic miniature photography feel.
OUTPUT STYLE
Premium stop-motion animation frame.
Absurdist claymation artwork.
Handcrafted miniature universe.
Aardman-inspired character design.
Museum-quality clay sculpture aesthetic.
Highly tactile materials.
Hyper-detailed clay textures.
Professional stop-motion production quality.
8K detail.
FINAL RESULT
The final image should feel like a frame from a big-budget stop-motion clay animation film, preserving the exact structure of the original image while transforming every element into a humorous, handcrafted, imperfect clay world full of personality and physical texture.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: 3D characters & caricatures
Full prompt, verbatim · 2,124 characters
Show the full prompt
Create a premium, highly recognizable brand mascot character designed for modern digital branding. The mascot should have a friendly, confident, energetic personality, a distinctive silhouette, expressive face, polished proportions, and a memorable visual identity. Make the character feel professional enough for a major technology, fintech, social-media, or digital-services brand while remaining approachable and fun.
FORMAT LOCK
Aspect ratio 3:4. Centered full-body mascot composition. Character clearly separated from the background with balanced negative space. Clean commercial presentation suitable for logos, websites, social media, advertisements, stickers, and brand merchandise.
IDENTITY LOCK
Create one consistent original mascot with distinctive facial features, signature hairstyle/head shape, recognizable clothing, accessories, and color identity. Keep the character design consistent throughout the image. Do not imitate an existing copyrighted mascot or recognizable character.
CHARACTER
Cute but premium 3D mascot with expressive eyes, confident friendly expression, slightly oversized head, polished stylized proportions, dynamic welcoming pose, clean modern outfit, subtle brand-inspired details, smooth surfaces, high-quality character modeling, appealing personality.
ENVIRONMENT
Minimal premium studio environment with a clean uncluttered background. Subtle floor shadow and soft atmospheric depth. No unnecessary objects competing with the mascot.
LIGHTING & STYLE
High-end 3D commercial character render, cinematic studio lighting, soft key light, subtle rim lighting, realistic reflections, polished materials, detailed textures, premium advertising aesthetic, sharp focus, clean edges, sophisticated color grading, professional brand-campaign quality.
NEGATIVE PROMPT
Low quality, blurry, distorted anatomy, extra limbs, extra fingers, malformed hands, asymmetrical eyes, duplicated features, messy background, excessive text, watermark, logo imitation, copyrighted character, generic stock mascot, flat lighting, oversaturated colors, noisy image, cropped character, character cut off at edges.
Figure potential: high (styled as a physical toy/figurine)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Toys & collectibles
Visible excerpt, verbatim · Copy gives these lines only
Create a premium handcrafted miniature diorama of {{Destination Landmark}} designed as a collectible glossy clay-and-resin sculpture.
FORMAT LOCK
Vertical 3:4 composition. Centered composition. Premium travel poster aesthetic. The landmark should be the clear focal point with a balanced miniature-world layout.
SCENE
Same text as the visible prompt except placeholder notation: site uses {{Destination Landmark}}, source uses [DESTINATION / LANDMARK]
Show the variant
Create a premium handcrafted miniature diorama of [DESTINATION / LANDMARK] designed as a collectible glossy clay-and-resin sculpture.
FORMAT LOCK
Vertical 3:4 composition. Centered composition. Premium travel poster aesthetic. The landmark should be the clear focal point with a balanced miniature-world layout.
SCENE
Recreate [DESTINATION / LANDMARK] as an ultra-detailed handcrafted miniature world. Place the iconic landmark at the center, surrounded by authentic local landscapes, vegetation, cultural elements, terrain, and destination-specific architecture. Add tiny realistic travelers naturally exploring the scene to enhance the miniature scale.
ENVIRONMENT
Include native plants, trees, rocks, mountains, water, pathways, or coastline appropriate to the destination. Sculpt thick impasto-style clouds and layered background scenery with handcrafted depth while maintaining a clean, elegant composition.
MATERIALS
Everything should appear handcrafted from glossy clay and polished resin with smooth ceramic surfaces, rounded organic shapes, subtle fingerprints, sculpted textures, realistic reflections, and premium collectible craftsmanship.
LIGHTING & STYLE
Soft golden-hour lighting. Tilt-shift photography effect. Shallow depth of field. Hyper-realistic 3D rendering. Vibrant destination-inspired colors. Luxury editorial travel poster aesthetic. Museum-quality collectible miniature. Premium Instagram-worthy composition. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid low quality, blur, flat illustrations, anime, cartoons, low-poly models, realistic full-scale photography, text, logos, watermarks, oversaturated colors, distorted perspective, duplicate landmarks, deformed miniature people, noisy textures, harsh shadows, washed-out colors, extra limbs, cropped composition, empty foreground, or cluttered layouts.
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 2,329 characters
Show the full prompt
Transform the uploaded person into The Sims 4 universe while preserving the exact identity, pose, composition, and environment of the original image.
FORMAT LOCK
Preserve the original aspect ratio, composition, framing, camera angle, crop, subject placement, background layout, and perspective. Do not zoom, reposition, redesign, or alter the scene structure.
IDENTITY LOCK — NON-NEGOTIABLE
Use the uploaded image as the only identity source. Preserve exact facial structure, facial proportions, hairstyle, hair color, skin tone, expression, body proportions, clothing, and recognizable identity traits. The person must remain 100% recognizable.
SIMS 4 TRANSFORMATION
Convert the subject into authentic Sims 4 character styling. Smooth slightly plastic skin, soft clean shading, simplified facial details, bright expressive eyes, polished game-engine rendering, clean Sims-style hair geometry, vibrant colors, and subtle stylization while maintaining realistic proportions.
Do not create a cartoon, chibi, Pixar character, or exaggerated caricature.
BACKGROUND TRANSFORMATION
Keep the exact original background, environment, objects, architecture, and scene layout unchanged, but transform everything into The Sims 4 visual style. Clean textures, bright in-game lighting, simplified materials, polished game aesthetics, and subtle game-world realism.
SIMS DETAILS
Add the iconic green Sims plumbob floating above the subject's head. The plumbob should be clean, centered, recognizable, and naturally integrated into the scene.
LIGHTING
Bright Sims-style daylight, soft shadows, even illumination, clean highlights, polished game-engine atmosphere, cheerful and realistic in-game lighting.
OUTPUT STYLE
Authentic Sims 4 screenshot aesthetic. Premium game-engine rendering. Clean textures. Bright colors. Smooth materials. High-resolution character design. Real person transformed into The Sims 4 world. Professional gameplay screenshot quality.
NEGATIVE PROMPT
Avoid changing identity, changing pose, changing clothing, adding accessories, changing hairstyle, changing camera angle, changing composition, realistic skin pores, noise, blur, CGI artifacts, cartoon style, Pixar style, chibi style, low quality, text, watermark, duplicated subjects, extra limbs, or background replacement.
Make the aspect ratio 3:4
Figure potential: medium (stylized character, needs 2D-to-3D interpretation)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Doodles on real photos
Full prompt, verbatim · 3,674 characters
Show the full prompt
Create an authentic South Park version of the uploaded person and seamlessly integrate it into the original photograph.
SCENE PRESERVATION — NON-NEGOTIABLE
Preserve the original image exactly, including the original background, environment, camera angle, perspective, framing, lighting, shadows, subject placement, and overall composition. Do not crop, zoom, reposition, replace, simplify, or redesign the original scene.
COMPOSITION — CRITICAL
Keep the original person exactly where they are in the photograph. Add an authentic South Park version of the same person into the existing scene. Analyze the available space and place the South Park character in the most natural and visually balanced position near the original person. The South Park character may stand beside, slightly in front of, or slightly behind the original person depending on the composition. The final image should look like the original photograph with a South Park companion version added into the scene. Do not create a split-screen layout, comparison poster, or replace the real person.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve only the essential identifying traits for the South Park version: hairstyle silhouette, hair color, basic face shape, clothing colors, recognizable visual identity, body posture, and any held objects. Simplify everything naturally to fit the South Park universe.
SOUTH PARK PROPORTION SYSTEM — CRITICAL
Create an extremely exaggerated authentic South Park character. Head occupies approximately 70% of total character height, body is extremely short and minimal, arms and legs are very small and stubby. The head should dominate the entire silhouette and appear larger than standard South Park proportions. The body should function almost entirely as support for the oversized head. The South Park character should be approximately 50–60% of the original person's height.
FACE DESIGN
Authentic South Park facial construction, simple oval eyes, small pupils, minimal closed mouth, flattened simplified nose, neutral expression, no realism, no depth, no detailed facial anatomy, no realistic shading.
POSE
The South Park version should mimic the original body position, arm placement, posture, expression, and any objects being held. Maintain the same overall character attitude and stance.
COLOR SYSTEM
Flat colors only, no gradients, lighting, shadows, highlights, outlines, black strokes, or comic linework. Use clean color-block separation only.
TEXTURE RULES
Skin must be completely flat color with no skin texture, grain, realism, or paper texture. Very subtle felt or paper-cut texture allowed only on clothing. The overall appearance should match classic South Park cutout animation.
STYLE
Authentic South Park television animation style, paper-cut animation aesthetic, simple geometric construction, clean vector-like appearance, minimalist character design, official TV-show quality.
OUTPUT STYLE
Original photograph preserved. Authentic South Park companion character naturally integrated into the scene. Official South Park character design, clean cutout animation look, sharp edges, high-resolution PNG appearance.
NEGATIVE PROMPT
Avoid replacing the original person, altering the original composition, split-screen layouts, comparison posters, realism on the South Park character, realistic skin, realistic eyes, detailed facial features, gradients, shadows, lighting effects, black outlines, comic strokes, painterly rendering, 3D rendering, Pixar style, anime style, chibi style, oversized muscles, long limbs, extra fingers, extra limbs, duplicated characters, text, watermark, cropped subjects, or background redesign.
Figure potential: medium (stylized character, needs 2D-to-3D interpretation)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Complete prompt, verbatim · 2,895 characters
Show the full prompt
Transform the uploaded image into the official Grand Theft Auto V promotional artwork style while preserving the exact original composition, framing, camera angle, crop, perspective, subject placement, pose, facial expression, clothing, environment, lighting direction, and scene structure.
FORMAT LOCK
Preserve the original aspect ratio, framing, crop, camera distance, perspective, subject scale, and environmental layout. Do not move, remove, reposition, replace, redesign, crop, zoom, or alter any element of the original image.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial structure, hairstyle, hair color, eyebrows, eyes, nose, lips, jawline, facial hair, skin tone, age appearance, and recognizable identity traits. The subject must remain immediately recognizable.
COMPOSITION LOCK — CRITICAL
Maintain the exact original pose, body posture, arm positions, hand positions, head tilt, gaze direction, facial expression, camera angle, perspective, and framing. The final image should match the original photograph exactly, with only the visual style transformed.
GTA V POSTER TRANSFORMATION
Convert the entire image into authentic Rockstar Games promotional artwork. Official GTA V loading-screen illustration style, promotional poster aesthetic, stylized realism, premium digital painting quality, recognizable GTA V visual identity.
ART STYLE
Bold black outlines, clean vector-like surfaces, semi-realistic painted textures, comic-book-inspired shading, crisp shadow shapes, smooth color transitions, stylized but recognizable facial rendering, polished promotional artwork finish.
COLOR GRADING
Vibrant GTA V color palette, rich contrast, saturated highlights, clean shadows, Rockstar-style color treatment, premium promotional artwork grading.
ENVIRONMENT
Transform the environment into GTA V promotional artwork while preserving the exact original scene layout, perspective, lighting direction, and object placement.
LIGHTING
Maintain the original lighting direction and mood while rendering it in GTA V promotional illustration style. Clean painted highlights, crisp shadow rendering, stylized Rockstar artwork lighting.
STYLE
Official GTA V loading-screen artwork, Rockstar Games promotional illustration, premium video-game poster art, comic-book realism, stylized digital painting, iconic GTA V advertising aesthetic.
OUTPUT STYLE
Authentic GTA V promotional poster, Rockstar Games loading-screen artwork, high-detail illustration, premium digital painting, vibrant stylized realism, ultra-detailed 8K quality.
Negative prompt:
Avoid photorealistic photography, Unreal Engine rendering, MetaHuman style, anime style, cartoon style, low-poly graphics, 3D CGI appearance, realistic skin pores, painterly brush chaos, text, logos, watermarks, HUD elements, loading screens, subtitles, altered composition, or modified subject placement.
Figure potential: medium (stylized character, needs 2D-to-3D interpretation)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 3,941 characters
Show the full prompt
Transform the uploaded person into the central protagonist of an original 1990s West Coast open-world crime-game cover poster. Create a premium illustrated game-cover collage with a dominant hero portrait surrounded by layered cinematic Los Santos street scenes, lowriders, palm trees, neon signs, graffiti, rooftops, city skylines, and helicopters.
FORMAT LOCK
Vertical 3:4.
Full-frame premium game-cover artwork.
Asymmetrical layered composition.
Overlapping scenes and angled borders.
No regular six-panel grid.
Strong depth, dramatic perspective, dense but organized layout.
Large central hero portrait with smaller story scenes around it.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the exact face, bone structure, eyes, nose, lips, jawline, skin tone, freckles, hairstyle, hairline, proportions, and recognizable identity.
The same person must appear consistently in every scene. No generic replacement or facial redesign.
ART STYLE
Original hand-painted 1990s West Coast video-game promotional art combining comic-book illustration, airbrushed poster art, cinematic character painting, retro magazine artwork, gritty street photography, and distressed screen-print textures.
Use bold shapes, painterly highlights, black outlines, halftone grain, aged paper, dramatic shadows, and saturated retro colors.
CHARACTER
Make the protagonist dominant and confident.
Hero: waist-up portrait wearing a black varsity jacket, vintage graphic T-shirt, baggy 1990s trousers, sneakers, gold hoops, layered gold chains, and subtle jewelry.
Surround her with six cinematic moments:
LOWRIDER — leaning against a chrome classic lowrider on a palm-lined sunset boulevard.
MOTEL — standing beneath glowing vintage neon at night.
ARCADE — sitting outside a retro arcade/record store.
STREET — walking through a busy 1990s Los Santos neighborhood.
ROOFTOP — overlooking the glowing skyline with a distant police helicopter.
NIGHT DRIVE — standing beside a lowrider with red taillights and downtown lights behind her.
ENVIRONMENT
1990s Southern California atmosphere: palm trees, telephone poles, overhead wires, old sedans, lowriders, neon storefronts, graffiti, payphones, arcades, record stores, rooftops, highways, streetlights, hazy mountains, and downtown skyscrapers.
Add subtle fictional signs such as LOS SANTOS, WEST COAST, GROVE STREET, SUNSET BLVD, PACIFIC AVE.
TITLE DESIGN
Large lower-center vintage game-cover typography:
GRAND THEFT AUTO
with oversized white letters, thick black outline, dimensional shadow, and distressed print texture.
Underneath:
SAN ANDREAS
in dramatic gothic/blackletter styling.
Keep the title extremely readable and integrated into the artwork.
GRAPHIC DETAILS
Add subtle vintage cover elements: rating badge, barcode, promotional sticker, fictional publisher mark, tiny legal text, map fragments, graffiti marks, worn edges, halftone texture, and distressed ink.
Do not include a reference-photo card or inset reference image.
LIGHTING & STYLE
Golden-hour scenes: orange sunlight, gold rim light, long shadows, warm reflections, atmospheric haze.
Night scenes: deep blue shadows, neon red, amber headlights, glowing skyscrapers, cinematic rim lighting.
Overall palette: burnt orange, sunset gold, faded teal, burgundy, cream, charcoal, muted green, dark navy.
Premium, gritty, cinematic, nostalgic, collectible 1990s poster finish.
NEGATIVE PROMPT
No identity drift, different person, generic face, altered facial features, changed eyes, nose, lips, jawline, hairstyle, inconsistent identity, distorted anatomy, malformed hands, extra fingers, warped vehicles, modern clothing, smartphones, futuristic elements, modern architecture, symmetrical six-panel grid, repetitive scenes, flat composition, excessive photorealism, glossy modern advertising, garbled text, misspelled signs, excessive text, reference-photo inset, watermark, low resolution, washed-out colors, excessive blur.
Figure potential: medium (stylized character, needs 2D-to-3D interpretation)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 2,360 characters
Show the full prompt
Transform the uploaded image into clean high-quality retro pixel art while preserving the original composition, subject placement, camera angle, proportions, pose, clothing, and overall scene structure.
FORMAT LOCK — NON-NEGOTIABLE
Preserve the original image aspect ratio.
Preserve the original composition and framing.
Do not crop, zoom, reposition, or redesign the scene.
Maintain the original visual hierarchy and subject placement.
SUBJECT PRESERVATION
Preserve the main subject, recognizable silhouette, facial direction, clothing structure, pose, body proportions, and key visual features.
Reduce unnecessary detail while keeping the subject immediately recognizable.
Do not alter the person, anatomy, outfit, or scene structure.
PIXEL ART STYLE
Convert the entire image into premium 16-bit retro pixel art.
Use a strict low-resolution pixel grid across the entire image.
Downsample the image into large visible pixels before rendering.
Maintain consistent pixel size throughout the artwork.
Create clean pixel-perfect geometry.
Use crisp hard edges.
Avoid semi-realistic rendering.
COLOR SYSTEM
Use a limited retro gaming color palette.
Strong contrast between shapes.
Clear color separation.
Carefully selected colors with excellent readability.
Avoid excessive color variation.
SHADING
Use simplified retro-game shading.
Flat color regions.
One to two shading tones per surface.
Minimal highlight usage.
No realistic gradients.
No soft lighting transitions.
ENVIRONMENT
Convert all environmental elements into pixel-art equivalents.
Background objects, props, architecture, vegetation, terrain, sky, and lighting should follow the same pixel-art language.
Keep scene depth and perspective while simplifying details.
VISUAL QUALITY
Authentic retro game aesthetic.
Clean sprite-art readability.
Consistent pixel density.
Balanced visual design.
No blur.
No anti-aliasing.
No smooth transitions.
No noise.
No artifacts.
No smudging.
No compression issues.
OUTPUT STYLE
Premium retro pixel art.
Classic 16-bit console game aesthetic.
Pixel-perfect rendering.
High readability.
Clean silhouette design.
Authentic game-screenshot appearance.
Crisp retro visuals.
Professional pixel-art craftsmanship.
The final image should feel like a polished SNES-era retro game screenshot while remaining faithful to the original image composition and subject.
Figure potential: medium (stylized character, needs 2D-to-3D interpretation)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 1,747 characters
Show the full prompt
Transform the uploaded image into a premium 16-bit retro video-game pixel-art illustration while preserving the original image exactly.
FORMAT LOCK
Vertical 3:4 composition. Preserve the exact original framing, subject placement, pose, proportions, camera angle, background layout, lighting direction, and overall composition.
STYLE TRANSFER
Apply only the pixel-art transformation. Convert the entire image into clean flat pixel blocks with chunky pixelated edges, limited saturated colors, crisp hard-edged shapes, visible pixel clusters, simplified details, and authentic 16-bit game aesthetics. Preserve the original visual structure and storytelling.
DETAIL & COLOR
Use a carefully controlled saturated retro-game palette with clear color separation, pixel-based highlights and shadows, simplified textures, and consistent pixel scale throughout the image. Keep important facial, clothing, environmental, and object details recognizable within the pixel-art language.
COMPOSITION LOCK
Do not move, add, remove, resize, replace, or redesign anything. The result should look like the exact same image has been completely redrawn using 16-bit pixels.
STYLE
Classic 16-bit console game aesthetic. Handcrafted pixel clusters. Crisp pixel edges. Retro RPG artwork. Clean sprite-like rendering. Nostalgic game visual style. Detailed but intentionally low-resolution appearance.
NEGATIVE PROMPT
Avoid changing the subject, pose, composition, camera angle, clothing, background, objects, or lighting structure. Avoid photorealism, smooth digital painting, 3D CGI, blurry pixels, anti-aliased edges, excessive gradients, modern vector graphics, added objects, removed objects, distorted anatomy, text, logos, watermarks, or low-quality pixel rendering.
Everything else on the site: drawn and painted portraits, doodles, paper, fashion, photo looks and posters. These restyle a photo but do not make a figure.
5 examples are not shown here because they show real public figures. See all on the site ↗
You + anime character streetwear poster
Figure potential: medium (character design, 2D source)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 1,825 characters
Show the full prompt
Create a premium Japanese streetwear campaign poster featuring the uploaded person together with {{Anime Character}} from {{Anime Series}}.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 composition.
Premium editorial campaign poster.
Magazine-cover quality.
No landscape, square, widescreen, or cropped layouts.
IDENTITY LOCK
Use the uploaded image as the only identity source.
Preserve exact facial structure, facial proportions, hairstyle, skin tone, age appearance, body proportions, and all recognizable identity traits.
The person must remain clearly recognizable.
CHARACTER LOCK
Preserve all iconic attributes of {{Anime Character}} including signature hairstyle, accessories, weapons, powers, clothing elements, and silhouette.
DUO CONCEPT
The uploaded person and {{Anime Character}} appear as elite partners within the same futuristic anime-techwear universe.
Strong chemistry.
Confident body language.
Premium promotional poster energy.
TECHWEAR STYLE
Oversized tactical jackets.
Utility straps.
Buckles.
Cargo pants.
Layered streetwear.
Technical fabrics.
Futuristic accessories.
Color palette inspired by {{Anime Character}}.
GRAPHIC DESIGN SYSTEM
Clean white editorial background.
Large bold typography:
{{Anime Character}} × {{User Name}}
Japanese typography.
Barcodes.
Technical data blocks.
Streetwear labels.
Editorial grids.
Graphic symbols.
Campaign branding elements.
Typography and graphic design should be as important as the characters.
RENDERING
Official anime promotional artwork quality.
Sharp cel-shading.
Ultra-detailed linework.
Modern anime illustration.
Luxury campaign design.
High contrast.
Dynamic perspective.
OUTPUT STYLE
Premium anime collaboration campaign.
Japanese streetwear poster.
Graphic-design-first composition.
Luxury editorial branding.
Official anime key visual quality.
8K detail.
Figure potential: medium (character design, 2D source)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,676 characters
Show the full prompt
Create a premium Japanese streetwear campaign poster featuring {{Anime Character}} from {{Anime Series}} reimagined as a luxury urban techwear fashion icon.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 composition.
Full-body hero shot.
Premium fashion campaign poster.
Magazine-cover quality.
Character centered.
Do not crop the feet, skateboard, or backpack.
CHARACTER DESIGN
Preserve all iconic features of {{Anime Character}}.
Recognizable hairstyle.
Recognizable facial features.
Signature weapons.
Signature accessories.
Signature personality.
The character must remain immediately recognizable.
STREETWEAR SYSTEM
Luxury Japanese techwear aesthetic.
Oversized bomber jacket.
Utility straps.
Hanging tags.
Buckles.
Technical fabrics.
Cargo pants.
Layered tactical clothing.
Streetwear patches.
Character-themed branding.
Premium sneakers.
Large utility backpack covered with custom patches and symbols inspired by the anime universe.
COLOR SYSTEM
Primary color:
{{Theme Color}}
Use monochromatic color coordination across:
Jacket.
Pants.
Shoes.
Backpack.
Typography.
Poster graphics.
Neon lights.
ENVIRONMENT
Luxury Tokyo streetwear campaign environment.
Character standing confidently on a skateboard.
Large illuminated poster panel directly behind the character.
Huge circular graphic behind the character's body.
Japanese urban fashion-store atmosphere.
Premium sneaker-store energy.
Subtle industrial walls.
Clean floor.
No crowded background.
POSTER DESIGN SYSTEM
Massive Japanese typography.
Vertical Japanese text columns.
Character name integrated into the design.
Small technical information blocks.
Barcodes.
Streetwear labels.
Graphic symbols.
Brand-style patches.
Magazine layout elements.
Editorial design hierarchy.
Japanese campaign poster aesthetic.
LIGHTING
Cinematic fashion photography.
Soft directional lighting.
Subtle neon reflections.
Premium retail-display lighting.
High contrast.
Luxury campaign mood.
RENDERING
Official anime promotional artwork quality.
Hyper-detailed illustration.
Premium character design.
Ultra-detailed clothing materials.
Sharp linework.
Luxury fashion rendering.
Modern anime-meets-streetwear aesthetic.
OUTPUT STYLE
Japanese streetwear campaign poster.
Anime fashion collaboration.
Luxury sneaker campaign.
Tokyo urban culture aesthetic.
Character branding poster.
Premium editorial design.
Graphic-design-first composition.
High-end collectible poster.
Ultra-detailed 8K quality.
The final image should feel like an official Tokyo streetwear collaboration campaign featuring {{Anime Character}} as the face of a luxury Japanese fashion brand.
Make the aspect ratio 3:4
Figure potential: medium (character design, 2D source)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 3,500 characters
Show the full prompt
Use the uploaded photograph as loose visual inspiration and transform its fashion attitude, pose energy, and urban atmosphere into a completely original premium Japanese anime/manga illustration. Do not reproduce the photograph exactly.
FORMAT LOCK
Vertical 3:4 composition. Dynamic seated portrait. Redesign the framing, camera perspective, character placement, environment, lighting, colors, and object arrangement while maintaining a stylish urban-anime feeling.
CHARACTER DESIGN
Create a completely original fictional anime character inspired only by the general fashion attitude of the reference. Design new facial features, hairstyle, face shape, clothing, sunglasses, hat, jewelry, shoes, patterns, and accessories. Do not preserve distinctive real-person identity characteristics.
SCENE
Create a fictional colorful urban street with redesigned architecture, storefronts, vehicles, street furniture, signs, windows, posters, and environmental details. All businesses, logos, advertisements, newspaper names, and signage must be fictional and newly designed.
POSE & COMPOSITION
Take loose inspiration from the reference’s seated pose and upward-looking camera energy, but create a new natural anime pose. Adjust the arms, legs, posture, camera angle, and framing enough to make the composition original.
CLOTHING
Create an original sporty streetwear outfit with an oversized jersey, loose trousers, fashionable sneakers or boots, original sunglasses, a redesigned hat, and unique accessories. Use fictional patterns, numbers, labels, and graphics.
ANIME STYLE
Premium cinematic Japanese anime/manga aesthetic. Bold black ink outlines, expressive contours, refined anime facial construction, detailed manga linework, sophisticated cel shading, crisp highlights, detailed hair, dynamic fabric folds, dimensional anatomy, vibrant controlled colors, cinematic contrast, and richly illustrated environments.
LIGHTING & COLOR
Use a vivid but sophisticated urban palette with natural daylight, anime-style highlights, controlled cel shadows, subtle reflected light, atmospheric depth, and cinematic contrast.
ENVIRONMENTAL STORYTELLING
Build a lively fictional city street with its own visual identity. Include fictional store names, street signs, newspaper titles, advertisements, vehicle markings, posters, and architectural details. Keep the background detailed but visually balanced.
FINAL LOOK
Stylish urban streetwear, confident fictional character, dynamic seated pose, colorful city environment, premium cinematic anime illustration. Ultra-detailed ink work, polished cel shading, expressive character design, cinematic depth, crisp environmental rendering, sophisticated color grading, professional high-resolution anime artwork.
NEGATIVE PROMPT
Avoid exact recreation of the reference, real-person identity, exact facial features, exact hairstyle, exact clothing, exact pose, exact camera framing, exact background layout, real-world logos, trademarks, advertisements, newspaper brands, copied signage, duplicate objects, extra characters, distorted anatomy, extra fingers, extra limbs, deformed hands, generic low-detail anime, 3D CGI, plastic skin, watermarks, or random recognizable brand names.
CREATIVE PRIORITY
The reference is inspiration only. Prioritize a substantially different fictional character, original clothing, original pose, original urban environment, fictional signage, and new visual details while preserving the broad feeling of stylish urban anime street photography.
6 examples are not shown here because they show real public figures. See all on the site ↗
Fan-art sketchbook page of any character
Figure potential: low (flat 2D look)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 3,507 characters
Show the full prompt
PROMPT SETTINGS
Character / Name: [WRITE CHARACTER NAME]
Outfit Color / Palette: [WRITE OUTFIT COLOR]
Text Language: [WRITE TEXT LANGUAGE]
Use the settings above for the entire prompt.
MAIN PROMPT
Draw the selected character as if an obsessed fan artist filled an entire sketchbook page with chaotic fandom energy.
Create a messy hand-drawn sketchbook illustration filled with overlapping full-body poses, tiny chibi doodles, exaggerated facial expressions, random close-ups of hands, eyes, accessories, outfit details, and expressive character moments.
FORMAT LOCK — NON-NEGOTIABLE
White background.
Full sketchbook page composition.
No grid.
No clean layout.
No structured panels.
No empty poster-style composition.
The page should feel completely filled with creative chaos.
CHARACTER IDENTITY
The character must be clearly recognizable as the selected character from the prompt settings.
Preserve the key facial features, hairstyle, silhouette, personality, signature styling, and recognizable traits of the character.
Do not redesign the character beyond the sketchbook fan-art style.
Do not make the character generic.
SKETCHBOOK COMPOSITION
Fill the page with multiple drawings of the same character.
Include messy overlapping full-body poses.
Add tiny chibi versions.
Add close-ups of eyes, hands, face details, accessories, shoes, hair, and outfit elements.
Use dynamic expressive poses.
Create the feeling of a fan artist obsessively drawing every detail of their favorite character.
The page should feel spontaneous, energetic, playful, emotional, and slightly chaotic.
FASHION STYLE
Style the character in aesthetic outfits using the selected outfit color / palette from the prompt settings.
Use layered fashion styling.
Include trendy outfit combinations.
Add expressive accessories.
Make the fashion feel stylish, modern, Pinterest-inspired, and character-specific.
Keep the outfits consistent with the personality of the selected character.
DECORATIVE ELEMENTS
Add flowers.
Foliage.
Stars.
Stickers.
Hearts.
Sparks.
Graffiti details.
Cool decorative shapes.
Small handwritten notes.
Fun fandom symbols.
Playful sticker-like accents.
All decorations should feel naturally drawn into the sketchbook page.
ART STYLE
Hand-drawn sketchbook illustration aesthetic.
Rough pencil sketches mixed with clean line art.
Colorful marker rendering.
Expressive brush strokes.
Anime-inspired doodles.
Modern fandom art style.
Trendy Pinterest art vibe.
Ultra-detailed 2D illustration.
Stylish creative chaos.
TYPOGRAPHY RULE
All visible text, notes, captions, labels, and handwritten details must be written only in the selected text language from the prompt settings.
Typography must be clean, readable, natural, and grammatically correct.
No broken letters.
No distorted symbols.
No random text.
No English text unless the selected text language is English.
OUTPUT STYLE
Chaotic fan sketchbook page.
Obsessed fan artist energy.
Expressive anime-inspired character doodles.
Layered fashion illustration.
White background.
No watermark.
High-detail 2D artwork.
Playful fandom composition.
Colorful marker and pencil sketch aesthetic.
NEGATIVE PROMPT
Avoid grid layouts, clean comic panels, empty composition, boring poses, generic character design, unreadable text, broken letters, random symbols, wrong language, English text when not requested, distorted anatomy, extra limbs, messy low-quality scribbles, watermark, logo, blurry details, low resolution, flat boring colors, and overly polished vector art.
8 examples are not shown here because they show real public figures. See all on the site ↗
Couple street-style iPhone lock screen
Figure potential: low (flat 2D look)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 3,646 characters
Show the full prompt
Create an ultra-realistic premium iPhone lock screen wallpaper featuring the referenced couple in a fashionable cinematic street-style composition.
FORMAT LOCK — NON-NEGOTIABLE
Vertical iPhone lock screen wallpaper format.
Full-screen mobile composition.
Optimized for premium iPhone wallpaper usage.
Do not generate square, landscape, cropped, poster, or desktop-style composition.
IDENTITY LOCK — NON-NEGOTIABLE
Use the uploaded reference image as the exact identity source for both people.
Fully preserve each person’s face, facial proportions, hairstyle, hair color, skin tone, body proportions, individuality, and recognizable features without changes.
Do not alter, beautify, redesign, stylize, age, or reinterpret their appearance.
Do not change face proportions.
Maintain a natural face ratio of 3:4.
Both people must remain clearly recognizable as the same individuals from the reference image.
SCENE & CAMERA
Ultra-realistic fashion couple shot from an extremely low-angle perspective, as if the camera is placed near the ground and looking upward.
The girl stands confidently in a stylish outfit while making a natural “V” gesture with her hand.
The guy walks confidently beside her with relaxed stylish body language.
Bright blue sky in the background.
Premium outdoor daylight look.
Fashion editorial energy.
Natural motion, natural body posture, and believable interaction.
Avoid stiff posing or artificial model behavior.
STICKER & DOODLE ELEMENTS
Add playful sticker-style decorative elements around the characters.
Include floral doodles, yellow dotted outlines, white sketch lines, hearts, stars, sparks, and small playful visual accents.
The doodles should feel clean, modern, fun, and intentionally designed.
Keep the sticker elements around the couple without covering or distorting their faces.
Do not overcrowd the composition.
Maintain a premium Apple Keynote-style visual balance.
IPHONE LOCK SCREEN INTERFACE
Add a realistic iPhone lock screen interface.
Include Dynamic Island at the top.
Include a small lock icon.
Include a large clock reading exactly: 08:30.
Include the date reading exactly: Friday, May 16.
Include battery charge reading exactly: 80%.
Include a clean weather widget.
Include realistic notification cards.
All interface text must be in English, fully legible, grammatically correct, clean, and free from errors.
No random symbols.
No AI text artifacts.
No misspelled words.
No broken numbers.
No corrupted UI elements.
CHIBI DETAIL
Place small chibi versions of the same couple next to the clock.
The chibi versions should be cute, clean, and playful while still resembling the couple.
Keep them small and decorative.
Do not make the main couple cartoonish.
Only the small chibi versions should be stylized.
LIGHTING & COLOR
Bright colorful daylight aesthetic.
Clean blue sky.
Premium high-contrast mobile wallpaper look.
Natural skin texture.
Realistic fabric details.
Sharp, vibrant, polished colors.
Avoid plastic skin, harsh overprocessing, or fake CGI lighting.
OUTPUT STYLE
Ultra-realistic iPhone lock screen wallpaper.
Premium Apple Keynote aesthetic.
8K quality.
Bright colors.
Fashionable couple editorial.
Playful sticker-doodle composition.
Clean mobile UI realism.
High-end social media wallpaper quality.
NEGATIVE PROMPT
Avoid unreadable text, English errors, random letters, distorted typography, plastic skin, distorted faces, changed identity, extra limbs, extra fingers, broken hands, unnatural anatomy, CGI look, full cartoon style, blurry details, low quality, watermark, logo artifacts, cropped faces, incorrect clock, incorrect date, incorrect battery percentage, and messy UI.
Prompt:
Create a hybrid photorealistic photo-and-mural composite using my uploaded face reference. Keep my facial identity exactly unchanged—same face shape, eyes, eyebrows, nose, lips, jawline, skin tone, hairline, hairstyle, and natural facial texture. No beautification, face replacement, or cartoonization.
Show me standing on a grey concrete sidewalk in front of a colorful cartoon street-art mural. I’m standing with my ankles crossed, one hand flat against the mural and the other in my jogger pocket, smiling warmly at the camera. I have long wavy brown hair with warm highlights.
Outfit: open oversized periwinkle-blue shirt, white cropped tank top, cream jogger pants with drawstring, white low-top sneakers, and a light-blue crossbody bag.
Inside the mural, create a large anime/semi-realistic illustrated version of me sitting on a rainbow, wearing the same outfit, winking and smiling. Use clean cel-shaded linework so the mural character clearly looks illustrated while the foreground figure remains fully photorealistic.
Add a playful mural background with a smiling cloud, paper airplanes, stars, flowers, pink balloon, purple butterfly, smiling sun, colorful rainbow, green hills, and small mountains.
Use soft natural daylight, realistic skin and fabric texture on the foreground figure, and vivid flat pastel colors with crisp linework in the mural. 35mm street-level photography, sharp realistic detail, natural colors, cinematic but authentic.
Cartoon Ghost Portrait + Sharp Streetwear Photo (Graffiti Wall)
CONCEPT — EXACT: A layered composite — a massive stylized cartoon/anime-style illustrated portrait fills the background against a vibrant graffiti-covered concrete wall, and a smaller sharp full-color photorealistic full-body portrait of the same woman stands in the foreground.
BACKGROUND CARTOON PORTRAIT — EXACT: Large flat-shaded anime/webtoon-style illustrated portrait, head tilted slightly, long straight dark hair falling past the waist, white rectangular sunglasses, glossy red lips, wearing an oversized yellow graphic t-shirt with a small "NY" logo patch and white sleeve stripes, rendered in clean cel-shaded illustration style with bold outlines, blended into the graffiti wall so it reads as a mural-style painting.
MAIN SUBJECT — FOREGROUND — EXACT: Full-body standing pose, hands tucked into the pockets of oversized cargo pants, head tilted slightly, calm confident expression, white rectangular sunglasses, long dark wavy hair with subtle highlights, thin silver necklace, small stud earrings, silver bracelet.
MAIN SUBJECT IDENTITY LOCK: Foreground photorealistic face must match the reference photo exactly — same jaw width, cheek fullness, eyebrow shape, eye shape/color, nose bridge, and lip shape. The background cartoon portrait may be stylized as illustrated art but should echo the same facial features and hairstyle.
BODY: Fair skin tone, slim build, relaxed confident standing posture.
OUTFIT: Oversized yellow crew-neck t-shirt with a small "NY" logo patch on the chest and white sleeve-cuff stripes, matching yellow cargo pants with multiple pockets and a drawstring waist, white sneakers, thin silver necklace, silver bracelet.
BACKGROUND/ENVIRONMENT — EXACT: A concrete wall densely covered in colorful graffiti street art — bold red, blue, orange, and yellow spray-paint shapes, splatter marks, and abstract tag lettering, a white vertical drainpipe running along the right edge, a strip of bright blue sky visible in the upper-right corner, plain grey concrete pavement beneath the subject's feet.
LIGHTING — EXACT: Bright natural daylight throughout, strong directional sunlight casting a defined shadow on the pavement, vivid clear outdoor lighting on the foreground figure.
COLOR GRADE — EXACT: Vivid saturated yellow dominant throughout both the outfit and the illustrated background portrait, bold multicolor graffiti tones providing contrast, natural warm skin tone on the photorealistic figure, clean vibrant streetwear-editorial color grading.
CAMERA: Shot on 35mm lens, natural street-level framing, sharp photorealistic focus on the foreground figure, background cartoon portrait cleanly illustrated with flat shading and bold linework.
Ultra photorealistic — zero AI artifacts, no watermark.
IMPORTANT — Use ONLY my uploaded face reference image for the facial structure and identity of the photorealistic foreground figure. My face must remain EXACTLY the same as my real appearance with zero identity change. Preserve my exact hairstyle, hairline, forehead shape, eyebrows, eye shape, nose structure, lips, jawline, skin tone, facial texture, and feminine proportions exactly as shown in the uploaded image. Do NOT beautify, cartoonize, smooth skin, or replace my face with a random celebrity or model face in the foreground portrait. Identity preservation is the highest priority for the photorealistic figure.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 1,794 characters
Show the full prompt
Transform the uploaded person into a gothic whimsical hand-drawn storybook character while preserving their recognizable identity, facial structure, hairstyle, skin tone, clothing, and overall likeness.
FORMAT LOCK
Vertical 3:4 composition. Full-body character scene. Clean illustrated composition with strong visual storytelling.
CHARACTER STYLE
The person and {{Animal}} share the same surreal illustrated universe. Both have oversized staring eyes, elongated spindly proportions, tiny mouths, bony awkward postures, and slightly melancholic yet charming expressions. Keep the human recognizable while applying the stylized proportions consistently.
DRAWING STYLE
Loose scratchy black ink linework. Rough sketch texture. Scribbled shading. Crooked perspective. Uneven hand-drawn contours. Naive expressive illustration. Slightly imperfect handmade appearance. No clean digital vector lines.
ENVIRONMENT
Create {{Scenery}} with warped architectural or natural forms, barren trees, strange silhouettes, and subtly distorted perspective. Keep the environment eerie but playful rather than frightening.
COLOR & ATMOSPHERE
Use {{Colors}} as the primary palette. Flat saturated color blocks combined with a softly glowing gradient sky. Gothic whimsical mood. Eerie-cute atmosphere. Surreal indie storybook aesthetic. Charming melancholy.
OUTPUT STYLE
Hand-drawn gothic storybook illustration. Surreal indie artwork. Expressive naive drawing style. Strong silhouettes. Premium illustrated-poster composition. Detailed ink texture. No realism.
NEGATIVE PROMPT
Avoid photorealism, realistic anatomy, clean vector graphics, glossy 3D rendering, anime style, overly polished digital painting, excessive horror, gore, text, logos, watermarks, blurry details, distorted identity, or cluttered composition.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,768 characters
Show the full prompt
Transform the uploaded person or character into a premium two-dimensional alternative character illustration while preserving the recognizable identity and key visual characteristics from the reference.
FORMAT LOCK
Vertical 3:4 composition. The character occupies approximately 70–75% of the frame. Waist-up framing. Slightly off-center composition with a subtle camera tilt. Premium editorial poster design.
CHARACTER DESIGN
Create [CHARACTER] in [POSE], with [HAIR] and [EXPRESSION]. Use an intentionally stylized face with an oversized rounded head, short lower face, narrow jaw, long thin neck, slightly enlarged ears, large pale oval eyes with tiny off-center pupils, thick horizontal black eyebrows, heavy upper eyelids, subtle mauve under-eye accents, a small broad triangular nose, and a wide asymmetrical grin with individually drawn irregular teeth.
Keep the character expressive, strange, playful, and slightly surreal without losing the recognizable qualities of the reference.
CLOTHING
Dress the character in [CLOTHING], simplified into broad graphic shapes with chunky seams, minimal folds, and a few small hand-drawn accessories.
ILLUSTRATION STYLE
Confident black-ink contours with thick outer silhouettes, medium-weight facial lines, thin angular hair strands, and sparse scratch marks around the eyes, hair, and clothing seams. Use flat opaque color fields, one hard-edged cel-shadow shape per form, sparse pencil hatching, subtle ink wobble, slight print misregistration, and restrained worn-paper texture.
COLOR PALETTE
Use a controlled [COLORS] palette. Keep every color clearly separated, moderately saturated, and visually balanced. Avoid gradients and excessive shading.
BACKGROUND
Place the character against [BACKGROUND], simplified into large flat graphic shapes with a few loosely drawn environmental details. Add oversized rough hand-painted lettering reading ”[TEXT]”. Keep typography partially cropped and integrated into the poster composition.
COMPOSITION
Off-center waist-up framing. Slight camera tilt. Character dominant in the frame. Cropped background typography. Loose poster margins. Lightly weathered printed edges. Strong graphic hierarchy and generous negative space.
STYLE
Contemporary alternative character illustration. Hand-drawn editorial poster. Experimental ink artwork. Retro print aesthetic. Modern graphic design. Flat cel-shaded illustration. Premium art-book quality. Expressive, quirky, stylish, and intentionally imperfect.
NEGATIVE PROMPT
Avoid photorealism, 3D CGI, realistic rendering, gradients, excessive shading, overly polished digital art, generic cartoon style, distorted anatomy, extra limbs, unreadable text, random lettering, excessive background detail, watermarks, logos, or low-quality rendering.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 1,664 characters
Show the full prompt
Transform the uploaded image into a premium contemporary naïve illustration while preserving the exact facial identity, facial structure, hairstyle, expression, body proportions, pose, clothing silhouette, composition, and overall scene from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Preserve the original framing, camera angle, perspective, subject placement, and overall composition. Clean centered layout with a premium poster aesthetic.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, expression, body proportions, clothing silhouette, and recognizable identity. Simplify the illustration without changing the person’s appearance.
ILLUSTRATION STYLE
Create a contemporary authorial naïve illustration with a minimalist hand-drawn aesthetic. Use playful proportions, slightly irregular hand-drawn outlines, expressive decorative linework, and intentionally imperfect craftsmanship. Simplify forms while keeping the subject immediately recognizable.
COLOR SYSTEM
Use flat high-contrast color blocks with a refined contemporary palette. No gradients. No shadows. No realistic lighting. No 3D effects. Clean graphic shapes with bold visual balance.
STYLE
Premium designer poster. Editorial illustration. Modern naïve art. Handmade graphic illustration. Museum-quality print design. Minimalist composition. Contemporary art-book aesthetic. Expressive handcrafted character. Ultra-clean premium artwork.
NEGATIVE PROMPT
Avoid photorealism, gradients, shadows, 3D effects, CGI appearance, digital painting, excessive detail, messy linework, blurry edges, distorted anatomy, extra limbs, text, logos, watermarks, or low-quality rendering.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 1,734 characters
Show the full prompt
Transform the uploaded portrait into a premium contemporary flat editorial illustration while preserving the person’s recognizable facial identity, hairstyle, skin tone, expression, and overall appearance.
FORMAT LOCK
Vertical 3:4 composition. Centered editorial portrait. Clean modern framing. Premium fashion-art poster aesthetic.
ILLUSTRATION STYLE
Refined semi-realistic character drawing combined with clean graphic shapes and elegant digital linework. Smooth facial anatomy, expressive eyes, simplified realistic features, confident contours, and carefully controlled stylized proportions. Keep the illustration polished, sophisticated, feminine, and contemporary.
COLOR & SHADING
Use flat color blocking with clean solid shapes, crisp edges, subtle tonal separation, and minimal texture. Render the hair as layered graphic masses with selective fine strands and flowing accents. Use softly stylized skin with simplified highlights and shadows. Keep clothing as clean flat fills with restrained folds and graphic shadow shapes.
BACKGROUND
Solid orange background covering the entire canvas edge to edge. Completely uniform and flat. No texture, pattern, scenery, objects, gradient, vignette, or background shadows.
STYLE
Modern fashion editorial illustration. Polished digital portrait. Contemporary graphic art. Clean, elegant, feminine, sophisticated, and highly refined. Premium illustration finish with simplified dimensionality and strong visual clarity.
NEGATIVE PROMPT
Avoid photorealism, 3D rendering, excessive texture, painterly brushwork, complex backgrounds, gradients, scenery, unnecessary details, heavy shadows, messy linework, distorted anatomy, extra features, text, logos, watermarks, or low-quality rendering.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,144 characters
Show the full prompt
Transform the uploaded person into a premium minimalist black-and-white editorial illustration while preserving the exact facial identity, facial structure, hairstyle, expression, body proportions, pose, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Pure white background. Centered composition with generous negative space. The character should occupy approximately 30–40% of the frame, creating a clean modern editorial layout.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial features, hairstyle, facial proportions, expression, clothing silhouette, and recognizable appearance. Simplify the illustration without changing the person’s identity.
CHARACTER DESIGN
Create a stylized 2D character with a slightly oversized head, simplified facial features, tiny dotted eyes, expressive eyebrows, a small nose, and a subtle friendly smile. Dress the character in casual modern streetwear with oversized silhouettes, chunky sneakers, minimal accessories, and a relaxed confident posture.
ILLUSTRATION STYLE
Draw using confident black ink lines with varied line weight, rounded contours, loose sketch-like imperfections, expressive hand-drawn strokes, and clean vector-inspired linework. Use only pure black and pure white with no gray tones, colors, gradients, or realistic shading. Clothing should use bold black shapes contrasted against white fabric with minimal folds and simple contour lines.
BACKGROUND
Completely blank white background with only a soft ground shadow or a subtle floor line beneath the feet. No additional objects or decorative elements.
STYLE
Minimalist editorial illustration. Modern indie webcomic aesthetic. Scandinavian editorial drawing style. Playful character design. Clean ink sketch. Premium magazine artwork. Museum-quality print. Charming handmade illustration. Ultra-sharp high-resolution rendering.
NEGATIVE PROMPT
Avoid photorealism, color, gradients, gray tones, complex shading, cluttered backgrounds, excessive detail, distorted anatomy, extra limbs, text, logos, watermarks, blurry linework, or low-quality rendering.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,116 characters
Show the full prompt
Create a premium high-contrast monochrome vector-art portrait using the uploaded image as the only identity reference.
FORMAT LOCK
Vertical 4:5 composition, tight shoulder-up portrait, centered framing, premium poster aesthetic, clean studio presentation.
IDENTITY LOCK
Preserve the exact facial structure, face shape, hairstyle, hairline, eyebrows, eyes, nose, lips, jawline, skin tone translated into monochrome, age appearance, and recognizable identity from the uploaded image, do not redesign or beautify facial features, the final artwork must immediately resemble the reference person.
PORTRAIT STYLE
Entire portrait rendered in high-contrast monochrome black and white, clean vector illustration, sharp cel-shading, bold graphic shapes, crisp edges, smooth gradients replaced with flat shadow planes, modern comic-book aesthetic, premium editorial vector style.
RIM LIGHTING
Strong vibrant red neon rim lighting outlining the hair, ears, jawline, neck, shoulders, and upper body, soft neon glow, realistic edge illumination, dramatic contrast against the black background.
WARDROBE
Open dark button-up shirt worn over a clean white t-shirt, minimalist square pendant chain necklace, modern fashion styling, simplified into clean monochrome vector shapes.
BACKGROUND
Pure matte black background, completely clean, no textures, no gradients, no objects, no distractions.
LIGHTING
Dramatic studio lighting, high contrast, bold shadow separation, cinematic edge lighting, premium editorial mood.
STYLE
Modern vector illustration, monochrome graphic portrait, luxury editorial poster, comic-book inspired artwork, clean geometric shapes, premium contemporary design.
OUTPUT STYLE
Premium vector portrait, monochrome black-and-white artwork, bold cel-shaded illustration, red neon rim lighting, gallery-quality poster, ultra-clean graphic design, ultra-detailed 8K quality.
NEGATIVE PROMPT
Photorealism, painterly style, watercolor, sketch, rough brush strokes, anime, low detail, noisy textures, blurry edges, colorful background, text, watermark, logos, distorted face, extra accessories, cluttered composition.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Doodles on real photos
Full prompt, verbatim · 2,177 characters
Show the full prompt
Preserve the original image exactly while adding a tasteful hand-drawn overlay on top of the photograph. Maintain the original composition, framing, camera angle, subject placement, lighting, colors, and scene structure.
FORMAT LOCK
Preserve the original aspect ratio, crop, perspective, subject scale, and image composition. Do not alter the underlying photograph.
HAND-DRAWN OVERLAY SYSTEM
Add thin {{Overlay Color}} hand-drawn lines as if sketched directly onto the photograph with a gel pen. Lines should feel spontaneous, imperfect, slightly shaky, and naturally hand-drawn.
Use a single-stroke doodle style with rough, uneven energy rather than polished illustration.
OUTLINE DETAILS
Create selective outlines around the outer edges of important subjects, objects, clothing details, accessories, or environmental elements. Do not fully trace everything. Keep the effect light and stylish.
GUIDE ELEMENTS
Add a few subtle arrows, dotted paths, circles, underlines, or sketch marks that naturally guide the viewer's eye through the composition.
HANDWRITTEN TEXT
Add short handwritten notes using
{{Text 1}}
{{Text 2}}
{{Text 3}}
Use a casual diary-like handwriting style. Text should feel personal, sweet, positive, and naturally integrated into the image.
DECORATIONS
Add a small amount of {{Decoration Style}} along with subtle doodles such as tiny stars, steam lines, smiley faces, floating accents, simple symbols, and playful sketch marks.
Use decorations sparingly and leave generous negative space.
STYLE
Modern scrapbook aesthetic, diary doodles, casual journal annotations, social-media sketch overlay, handwritten memory-book style, effortless and relaxed creative direction.
OUTPUT STYLE
Authentic photograph with elegant hand-drawn overlays, white pen doodles, casual diary annotations, minimal sketch elements, stylish personal-memory aesthetic, clean and premium visual balance.
NEGATIVE PROMPT
Avoid clutter, excessive doodles, large text blocks, colorful drawings, thick marker strokes, graffiti style, comic-book effects, heavy illustrations, full-image tracing, stickers, emojis, watermarks, excessive decorations, or overwhelming the original photograph.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Doodles on real photos
Full prompt, verbatim · 2,854 characters
Show the full prompt
Transform only the people in the uploaded photograph into adorable hand-drawn doodle characters while keeping the entire original photographic background unchanged.
FORMAT LOCK
Vertical 3:4 composition. Preserve the original framing, camera angle, perspective, lighting, colors, and scene structure.
BACKGROUND LOCK — CRITICAL
Keep the original sky, landscape, buildings, water, furniture, ground, plants, railings, objects, lighting, colors, perspective, textures, and environmental details fully photographic. Do not redraw, simplify, stylize, illustrate, or apply doodle effects to anything except the people.
PEOPLE TRANSFORMATION
Replace each person with a charming naive doodle character while preserving the exact number of people, original position, relative scale, orientation, head direction, body direction, pose, gestures, arm and leg positions, interactions, hairstyle, clothing colors, and major accessories.
If a person faces away, keep them back-facing. If sideways, keep them sideways. If facing forward, keep them forward-facing. Never rotate a person toward the camera or invent facial features that are not visible.
DOODLE STYLE
Adorable imperfect hand-drawn characters with oversized round heads, tiny compact bodies, short simplified limbs, tiny hands and feet, cute awkward proportions, loose scribbled hair, tiny dot eyes and simple facial features only when visible, and subtle rosy scribbled cheeks when appropriate.
DRAWING TECHNIQUE
Loose naive children’s-sketch aesthetic. Thin shaky black outlines, imperfect shapes, overlapping sketch lines, scribbled colored-pencil or crayon fills, uneven coloring, small white gaps, rough edges, and intentionally handmade imperfections. Keep the characters simple, charming, expressive, and slightly exaggerated.
INTEGRATION
Maintain accurate scale, ground contact, depth, perspective, lighting interaction, and natural occlusion. Doodle characters should occupy exactly the same locations as the original people and interact naturally with photographic objects. Objects being held or touched remain fully photographic.
COMPOSITION PRIORITY
Original photographic background
Exact person position and scale
Original body orientation
Original pose and gestures
Cute exaggerated doodle character design
FINAL LOOK
The image should feel like the real people were replaced with adorable hand-drawn doodle versions while the real-world photograph remained untouched. Create a convincing mixed-media contrast between realistic photography and handmade doodle characters.
NEGATIVE PROMPT
Avoid full-image illustration, background doodling, stylized scenery, realistic anatomy, anime, manga, 3D cartoon, polished digital art, vector lines, changed poses, changed orientation, invented faces, added people, altered background, altered objects, altered lighting, or unwanted environmental effects.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Doodles on real photos
Full prompt, verbatim · 2,219 characters
Show the full prompt
Create a premium editorial portrait featuring the uploaded person standing beside a playful ink-twin doodle version of themselves.
FORMAT LOCK
Vertical 4:5 composition. Full-body view. Both figures fully visible. Side-by-side layout. Clean minimalist composition.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the exact face, facial features, eye shape, nose, lips, skin tone, hairstyle, accessories, outfit, shoes, colors, and overall appearance. Do not invent or change any details.
REAL SUBJECT
Keep the person completely photorealistic with natural skin, realistic lighting, and authentic proportions. Match the outfit exactly from the reference. Give the subject a playful, slightly embarrassed, caught-in-the-moment expression with a subtle, relaxed pose.
INK TWIN
Create a rough black ink, charcoal, or marker doodle version of the same person directly on the wall beside them. The doodle should be recognizable only through its silhouette, hairstyle, outfit shape, and pose—not as a second realistic person. Use loose sketch lines, rough strokes, hand-drawn imperfections, and optional comic-style marks such as motion lines, stars, hearts, or scribbles.
POSE RULE
Generate a new playful or mischievous pose every time. The real subject performs a subtle version of the action, while the ink twin exaggerates the same movement with chaotic cartoon energy. Avoid pointing, finger guns, mirrored poses, repeated poses, or static body language.
BACKGROUND & LIGHTING
Clean white, cream, or light gray wall. Soft natural daylight-style lighting. Even illumination. Clean shadows. Minimal distractions.
STYLE
Luxury fashion editorial. Photorealism meets hand-drawn illustration. Contemporary magazine campaign. Minimalist storytelling. Premium mixed-media artwork. Modern social-media aesthetic.
OUTPUT
Ultra-detailed 8K quality. Sharp focus. Premium editorial photography. Clean composition. Fun, stylish, and highly shareable.
NEGATIVE PROMPT
Avoid invented appearance, outfit changes, realistic second person, flat shadow figures, colored cartoon characters, weak resemblance, clutter, stiff or mirrored poses, extra limbs, distorted hands, text, logos, watermarks, or low quality.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 1,748 characters
Show the full prompt
Transform the uploaded image into a premium minimalist hand-drawn doodle illustration while preserving the original composition, subject placement, pose, facial identity, body proportions, and overall scene structure.
FORMAT LOCK
Vertical 3:4 composition. Preserve the original framing, camera angle, perspective, and subject placement. Keep the composition clean, balanced, and centered.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, expression, body proportions, clothing silhouette, and recognizable identity. Simplify the design without changing the person’s appearance.
DOODLE STYLE
Convert the entire scene into a charming handmade doodle illustration using simple brush-pen outlines with slightly wobbly, imperfect strokes. Simplify every element into cute, childlike shapes while keeping the subject immediately recognizable. Use expressive minimal linework, playful proportions, and a naïve sketchbook aesthetic with natural handmade imperfections.
COLOR & TEXTURE
Clean white paper background. Soft watercolor ink washes. Light monochrome or pastel accents. Visible paper texture. Hand-drawn brush strokes. Minimal shading. Loose sketch lines. Organic imperfections. Simple cartoon illustration with a warm handcrafted feel.
STYLE
Minimalist doodle art. Children’s sketchbook illustration. Handmade ink drawing. Whimsical character design. Editorial illustration. Modern paper-art aesthetic. Cute, expressive, playful, and elegant. Ultra-detailed handcrafted quality.
NEGATIVE PROMPT
Avoid realistic rendering, digital painting, CGI appearance, excessive detail, sharp vector lines, perfect symmetry, cluttered composition, dark backgrounds, distorted anatomy, extra limbs, text, logos, watermarks, or low-quality linework.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Doodles on real photos
Full prompt, verbatim · 2,099 characters
Show the full prompt
Transform the uploaded image into a premium rooftop lifestyle editorial while preserving the exact facial identity, facial features, hairstyle, skin tone, body proportions, outfit, shoes, pose, camera angle, and overall composition from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Preserve the original framing, perspective, and subject placement. Premium Instagram editorial aesthetic.
IDENTITY LOCK
Preserve the exact facial structure, facial proportions, hairstyle, skin tone, expression, clothing, shoes, and recognizable identity. Do not redesign the face, replace the outfit, or alter the body proportions.
ENVIRONMENT
Place the subject confidently on a modern rooftop walkway during golden hour. A contemporary city skyline stretches across the background beneath a soft blue sky. Warm sunlight, shallow depth of field, and cinematic urban photography create a premium lifestyle atmosphere.
DOODLE EFFECT
Behind the subject, create large hand-drawn white angel wings perfectly aligned with the back as if painted directly onto the wall. Add subtle white sketch outlines around the clothing and body, along with small floating ink splashes, tiny stars, a simple halo above the head, and a few playful hand-drawn doodles. The illustrations should blend naturally with the real photograph without overpowering the subject.
LIGHTING & STYLE
Soft golden-hour sunlight. Cinematic contrast. Natural shadows. DSLR-quality realism. Crisp skin and fabric textures. Luxury fashion editorial. Modern Instagram aesthetic. Clean white line-art illustration. Premium color grading. Shallow depth of field. Ultra-detailed 8K rendering.
OUTPUT
Photorealistic rooftop fashion portrait enhanced with elegant hand-drawn doodles. Clean, playful, modern, and premium editorial quality.
NEGATIVE PROMPT
Avoid changing the face, hairstyle, outfit, shoes, colors, pose, camera angle, or composition. Avoid cartoon faces, anime style, distorted anatomy, extra limbs, duplicate features, blurry details, messy doodles, low quality, text, logos, watermarks, or unrealistic proportions.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials, Doodles on real photos
Full prompt, verbatim · 2,426 characters
Show the full prompt
Transform the uploaded image into a premium mixed-media lifestyle fashion editorial while preserving the exact facial identity, facial features, hairstyle, skin tone, body proportions, outfit, shoes, pose, camera angle, and overall composition from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Eye-level camera. Preserve the original framing, perspective, and subject placement. Premium Instagram editorial aesthetic.
IDENTITY LOCK
Preserve the exact facial structure, facial proportions, hairstyle, skin tone, expression, outfit, shoes, and recognizable identity. Do not redesign the face, replace the clothing, or alter the body proportions.
ENVIRONMENT
Place the subject naturally in front of a clean light-blue painted brick wall. Keep the composition minimal, balanced, and uncluttered with soft natural daylight and a modern editorial atmosphere.
DOODLE EFFECT
Behind the subject, create a glowing white hand-drawn line-art version of the same person in a different dynamic pose, raising one arm toward the sky. The doodle figure should appear as a luminous neon sketch with soft transparent strokes, glowing edges, sparkling particles, magical light effects, and a floating halo above the head. Blend the illustration seamlessly into the brick wall as if it were painted directly onto the surface.
DETAILS
Add thin white doodle outlines around selected parts of the real subject’s clothing and body. Include subtle glowing stars, sparkles, sketch accents, and playful illustration details that naturally connect the real subject with the doodle artwork without overwhelming the composition.
LIGHTING & STYLE
Soft natural daylight. Cinematic lighting. Premium mixed-media editorial photography. Glowing white doodle art. Clean modern Instagram aesthetic. Vibrant colors. Sharp focus. Ultra-detailed textures. Premium color grading. Magical yet realistic atmosphere. Ultra-detailed 8K rendering.
OUTPUT
Luxury fashion editorial blending photorealistic portrait photography with elegant glowing hand-drawn doodle illustrations. Clean, whimsical, modern, and premium visual storytelling.
NEGATIVE PROMPT
Avoid blurry details, low quality, extra limbs, duplicate body parts, distorted anatomy, bad hands, cropped feet, changed identity, altered outfit, deformed face, unrealistic proportions, oversaturated colors, noisy images, artifacts, text, logos, watermarks, or cluttered backgrounds.
Upload a reference image before generating.
MAIN PROMPT
Transform only the main subject in the uploaded photograph into a hand-drawn monochrome doodle sticker illustration, while keeping the entire real environment completely unchanged and photorealistic.
SUBJECT PRESERVATION
Preserve the subject’s exact pose, body proportions, facial expression, hairstyle, clothing, accessories, gesture, position, scale, and interaction with the scene. Recreate the existing appearance and outfit accurately without redesigning or simplifying important details.
DOODLE ILLUSTRATION STYLE
Use bold black ink outlines, simple off-white fills, minimal cross-hatching, subtle line variation, and slightly imperfect handmade strokes. Keep the illustration clean, editorial, playful, and authentically hand-drawn rather than digitally polished.
STICKER BORDER
Add a thick, irregular cream-white sticker border tightly following the illustrated subject’s silhouette. The border should feel naturally applied around the character while remaining clearly separated from the untouched photographic environment.
ENVIRONMENT LOCK
Keep all furniture, architecture, objects, background details, lighting, shadows, reflections, textures, camera angle, framing, depth, perspective, and spatial relationships exactly as in the original photograph. Do not stylize, redraw, blur, replace, or alter the real environment.
COMPOSITION LOCK
Keep the subject in the exact original position, scale, orientation, framing, and relationship to surrounding objects. Only the subject and its immediate sticker outline may become illustrated.
FINAL RESULT
The final image must look like a clean monochrome editorial doodle-sticker character seamlessly integrated into an otherwise completely untouched, realistic photograph.
NEGATIVE PROMPT
No pose change, no body reshaping, no clothing redesign, no hairstyle change, no expression change, no environment alteration, no background replacement, no new objects, no decorative doodles, no speech bubbles, no text, no captions, no logos, no extra characters, no cartoonized background, no painterly environment, no colorization, no excessive shading, no heavy cross-hatching, no glossy digital outlines, no distorted anatomy, no perspective changes, no blur, no loss of photographic realism.
FORMAT LOCK
Aspect ratio: 3:4.
Preserve the original photographic framing and composition.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 3,945 characters
Show the full prompt
PROMPT SETTINGS
Balloon Text: GOOD VIBE
Balloon Color: Metallic Orange
Scene Style: Preserve Original Reference Environment
Phone Color: Bright Yellow
Outfit Style: Preserve Original Outfit
MAIN PROMPT
Upload one reference image before generating.
Create a realistic mirror-selfie fashion photograph using the uploaded image as the identity, outfit, environment, and pose reference.
FORMAT LOCK — CRITICAL
Strict vertical 3:4 composition, full-body portrait, entire subject visible from head to shoes, portrait orientation only, no square crop, no 4:5 crop, no landscape composition, authentic smartphone mirror selfie, luxury editorial framing, Instagram fashion aesthetic, realistic smartphone-camera quality.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, hair color, sunglasses, body proportions, pose, expression, skin tone, outfit details, accessories, footwear, and overall recognizable appearance from the uploaded image. The final result must immediately resemble the reference person.
POSE LOCK — CRITICAL
Preserve the exact pose from the reference image, subject standing inside the original environment, maintain the exact body angle, stance, posture, head tilt, hand positions, phone position, gaze direction, and fashion attitude.
ENVIRONMENT LOCK — CRITICAL
Preserve the original environment exactly as shown in the reference image, preserve the original elevator architecture, reflective stainless-steel walls, elevator doors, elevator handrails, mirrors, reflections, floor materials, ceiling details, lighting fixtures, perspective, camera angle, camera distance, framing, composition, reflections, shadows, and overall atmosphere. Do not replace, redesign, modernize, relocate, simplify, or alter the environment in any way.
BALLOON INSTALLATION — CRITICAL
Fill the original environment with gigantic metallic foil balloon letters spelling "GOOD VIBE", each character must be an individual oversized foil balloon, letters should be dramatically larger than the subject, balloons completely dominate the space, some letters press against the walls, some letters touch the ceiling, some letters bend naturally due to limited space, some letters appear behind the subject, some letters appear close to the camera, some letters partially block architectural elements, the environment should feel almost too small to contain them.
BALLOON MATERIALS
Metallic pink foil balloons, highly reflective glossy finish, visible seams, inflation wrinkles, natural creases, premium foil texture, realistic air-pressure deformation, strong reflections from surrounding surfaces, photorealistic metallic appearance.
PHONE
Bright pink smartphone, natural mirror-selfie positioning, realistic reflections.
OUTFIT
Preserve the original outfit exactly from the reference image, preserve all clothing items, colors, accessories, bags, sunglasses, jewelry, footwear, textures, fabrics, and styling details. Do not alter, replace, redesign, or modify any wardrobe elements.
LIGHTING
Preserve the original lighting from the reference image, soft realistic highlights, natural reflections, realistic shadows, bright reflections on balloon surfaces, authentic smartphone-photo exposure.
STYLE
Hyper-realistic photography, luxury fashion campaign aesthetic, premium Instagram aesthetic, mirror-selfie realism, surreal oversized-object realism, ultra-detailed 8K quality.
OUTPUT STYLE
Photorealistic mirror selfie surrounded by giant metallic pink GOOD VIBE balloons, immersive oversized-balloon installation, realistic reflections, luxury fashion photography, premium social-media aesthetic.
NEGATIVE PROMPT
Square crop, landscape orientation, cartoon style, CGI appearance, matte balloons, floating letters, distorted anatomy, duplicated subject, altered outfit, altered environment, redesigned elevator, unrealistic lighting, text overlays, watermark, cropped feet, cropped head, incomplete body, extra people.
Make the aspect ratio 3:4
7 examples are not shown here because they show real public figures. See all on the site ↗
Mixed-media magazine poster portrait
Figure potential: low (flat 2D look)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,358 characters
Show the full prompt
Create a premium editorial mixed-media magazine poster featuring {{Character Name}} as a centered shoulder-up portrait, staring directly into the lens with a {{Expression}} expression.
FORMAT LOCK
Vertical 4:5 poster composition. Centered shoulder-up portrait. Black clothing extends to the absolute bottom edge. Large bold title at bottom. Clean vintage magazine cover layout.
CHARACTER STYLE
{{Character Name}} has {{Hair / Facial Hair}} and wears {{Black Clothing}}. High-contrast black ink portrait illustration, realistic facial structure, crisp linework, dense cross-hatching, retro stippling, sharp editorial shadows, monochrome ink drawing aesthetic.
EYE HIGHLIGHT
Add a slightly tilted translucent {{Highlight Color}} graphic bar strictly across the eyes. The bar must be see-through, like vellum with multiply-blend effect. Eyes, ink lines, facial details, and background text must remain visible through the color. Add subtle halftone dot texture and imperfect screen-print edges.
BACKGROUND
Cold ivory aged paper background, not yellow, not beige, not sepia. Dense tiny black editorial text columns fill the background. Clean justified newspaper grid, narrow gutters, negative space cutout around the head and shoulders. Add subtle printer registration marks, small barcode, serial number, crop marks, color test strip, and archival print details.
TYPOGRAPHY
At the bottom, over the black clothing, add huge bold white uppercase title: {{Character Name}}. Under it, add smaller widely spaced white subtitle: {{Subtitle}}. Typography must be clean, readable, heavy geometric sans-serif, premium magazine-cover style.
PRINT TEXTURE
Slight paper grain, light archive wear, subtle creases, small ink imperfections, clean black-and-white contrast, premium vintage editorial print quality. Keep the paper mostly off-white and neutral, never warm yellow.
OUTPUT STYLE
Avant-garde editorial magazine poster, monochrome ink portrait, translucent eye bar, dense text background, cold ivory paper, bold white typography, premium graphic design poster, museum-quality print aesthetic.
NEGATIVE PROMPT
Avoid yellow paper, sepia tone, warm beige background, muddy colors, opaque eye bar, hidden eyes, messy text, unreadable title, random symbols, weak typography, photorealism, cartoon style, glossy CGI, distorted face, extra limbs, watermark, logo artifacts.
7 examples are not shown here because they show real public figures. See all on the site ↗
Raw crayon and colored-pencil artwork
Figure potential: low (flat 2D look)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 1,695 characters
Show the full prompt
Transform the entire uploaded image into a raw crayon and colored-pencil mixed-media artwork while preserving the original composition, pose, framing, camera angle, subject placement, facial identity, clothing structure, hairstyle, background, and overall image layout.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 3:4 composition.
Preserve the entire subject within frame.
Do not generate landscape, square, widescreen, panoramic, or cropped compositions.
Maintain balanced margins around the subject.
Gallery-poster layout.
STYLE
Hand-drawn crayon and colored-pencil illustration. Visible wax crayon texture. Scribbled colored-pencil shading. Loose expressive strokes. Rough cross-hatching. Paper grain. Oil pastel smudges. Raw unfinished sketch energy. Childlike creativity mixed with professional art-school drawing style.
COLOR PALETTE
Use cobalt blue, cadmium yellow, vermilion red, burnt orange, warm beige, and deep graphite accents.
IMAGE PRESERVATION
Keep the same subject, same pose, same face direction, same body proportions, same clothing, same background structure, and same aspect ratio from the uploaded image. Do not redesign the person, change identity, replace the background, or create a new composition.
OUTPUT STYLE
Premium contemporary mixed-media portrait. Gallery-style hand-drawn artwork. Expressive crayon sketch aesthetic. Textured paper surface. Museum-quality art print. High-detail handcrafted illustration.
NEGATIVE PROMPT
Avoid photorealism, clean digital painting, polished cartoon style, vector graphics, smooth airbrush shading, plastic skin, changed identity, distorted face, extra facial features, text, watermark, logo, blurry details, and low-quality artifacts.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,104 characters
Show the full prompt
Transform the uploaded image into a funny childish crayon drawing that looks like it was made by a very young child around 3 to 4 years old.
REFERENCE PRESERVATION
Keep the basic idea, main subject, pose, composition, and important recognizable elements from the uploaded image.
The result should still loosely resemble the original image, but only in a simple, naive, childlike way.
Do not make it accurate, polished, realistic, or professionally illustrated.
CHILD DRAWING STYLE
Make the drawing look silly, clumsy, awkward, uneven, and adorably bad.
Use rough crayon strokes, scribbly outlines, wobbly shapes, uneven proportions, simple forms, random line pressure, imperfect coloring, messy fill areas, and childlike color choices.
The lines should not be straight, clean, smooth, or controlled.
FORMS & PROPORTIONS
Simplify the subject heavily.
Use incorrect proportions, oversized or tiny body parts, uneven eyes, crooked shapes, simple stick-like limbs when suitable, naive facial features, awkward anatomy, and a loose interpretation of the original image.
It should feel like a toddler tried to draw what they saw but only vaguely understood it.
COLOR & TEXTURE
Use simple crayon colors.
Visible waxy crayon texture, messy coloring outside the lines, flat childish colors.
No realistic shading, no gradients, no polished lighting, no adult-level color harmony.
COMPOSITION
Keep the original image layout roughly recognizable.
Allow the placement to feel slightly random and imperfect.
Avoid professional composition, realistic perspective, clean framing, or polished balance.
OUTPUT STYLE
Very young child’s crayon drawing.
Naive toddler art.
Silly and funny.
Messy and charming.
Awkward but recognizable.
Rough hand-drawn crayon texture.
Simple, uneven, playful, and imperfect.
NEGATIVE PROMPT
Avoid realistic proportions, accurate anatomy, clean line art, polished illustration, adult drawing skill, professional cartoon style, detailed shading, smooth gradients, perfect symmetry, clean composition, realistic lighting, digital painting, anime style, high-end illustration, or refined design.
Upload one reference image before generating.
MAIN PROMPT
Transform the uploaded image into a soft handcrafted paper-cut layered illustration while preserving the original composition, subject placement, pose, proportions, perspective, and recognizable features.
FORMAT LOCK
Vertical 3:4 composition. Preserve the original framing, camera angle, crop, and overall visual hierarchy.
STYLE
Soft papercraft diorama aesthetic with smooth rounded shapes, simplified cute proportions, minimal facial details, dot eyes, soft blush cheeks, and a warm storybook character design. Recreate the entire scene using stacked paper layers with visible dimensional depth, subtle shadows between layers, and clean cut edges resembling carefully crafted cardstock.
CHARACTER DESIGN
Keep the main characters recognizable while simplifying their forms into charming paper-cut figures. Add a distinct thick white outer paper layer around each main character, creating a clean sticker-like cut-paper border that physically separates them from the background. The white border must look like an intentional cardstock layer, not a glow or lighting effect.
PAPER MATERIAL
Use matte tactile materials resembling thick art paper or craft foam. Show subtle paper grain, layered edges, slight thickness, clean cuts, tiny handmade imperfections, and realistic contact shadows between overlapping layers.
COLOR PALETTE
Soft pastel colors with muted blues, gentle greens, warm neutrals, cream, and subtle complementary accents. Keep the palette balanced, calming, cozy, and harmonious.
LIGHTING
Soft diffused even lighting. Gentle ambient illumination. Subtle layer shadows. No harsh contrast. Enhance the physical depth of the paper layers while maintaining a soft handcrafted atmosphere.
COMPOSITION
Preserve the original scene structure and visual hierarchy. Keep characters, objects, and important environmental elements in their original positions while translating everything into the same paper-cut visual language. Maintain clear separation between foreground, midground, and background.
OUTPUT STYLE
Cozy handcrafted papercraft diorama. Modern children’s-book illustration. Decorative paper-cut artwork. Warm storybook aesthetic. Wholesome, gentle, playful, polished, tactile, and premium.
NEGATIVE PROMPT
Avoid photorealism, CGI, glossy surfaces, plastic appearance, neon colors, harsh lighting, excessive detail, realistic facial rendering, floating characters, glowing white outlines, flat digital illustration, messy edges, excessive shadows, distorted anatomy, changed composition, text, logos, watermarks, or low-quality rendering.
HOME KITCHEN ACCIDENT
Create an ultra-realistic accidental iPhone front-camera snapshot of an attractive young woman at home. The image should feel completely unplanned and accidentally captured while the phone was being picked up or moved near a kitchen counter.
The woman has a naturally attractive appearance, casual fitted homewear, slightly visible body shape, natural proportions, and effortless everyday beauty.
The framing should be awkward, slightly tilted, imperfect, and off-center. Part of the head may be cropped. Include subtle motion blur, smartphone noise, imperfect exposure, and front-camera softness.
The subject appears caught while reaching for something on the counter, turning around, opening a cabinet, or walking through the kitchen. She is not looking at the camera.
Realistic kitchen environment with cabinets, fridge, dishes, groceries, warm indoor lighting, and natural clutter.
The image should feel like a random accidental photo discovered later in the camera roll, not a selfie, not a photoshoot, and not social media content.
CAR PARKING SNAPSHOT
Create an ultra-realistic accidental smartphone selfie inside a parking garage.
An attractive young woman is captured unintentionally while sitting in a parked car. The phone appears to have slipped or moved while being picked up.
Low-angle front-camera perspective from lap level. Slight motion blur, imperfect crop, camera shake, off-center framing, and realistic smartphone noise.
The subject wears casual everyday clothing with a natural fitted silhouette. She looks away from the camera while adjusting her position or reaching for something.
Visible car interior details, dashboard, seat, window reflections, and parking garage lighting.
The image should feel completely accidental and authentic.
GROCERY STORE MOMENT
Create an ultra-realistic accidental front-camera photo captured inside a grocery store.
An attractive young woman is walking through the aisles when the phone accidentally captures an image from waist level.
Awkward low-angle framing, slight motion blur, rolling-shutter feel, imperfect crop, uneven lighting, and natural smartphone imperfections.
Store shelves, products, shopping carts, fluorescent lights, and customers appear softly blurred in the background.
The woman looks distracted and unaware of the camera. The image should feel mundane, random, and completely real.
ELEVATOR ACCIDENT
Create an ultra-realistic accidental smartphone front-camera image inside an elevator.
The phone captures the subject unintentionally while being moved or slipping from one hand to the other.
Attractive young woman wearing casual everyday clothing. Natural body shape and proportions.
Slight overexposure from elevator lights, awkward perspective, motion blur, partial crop, off-center composition, and realistic smartphone grain.
The subject appears to be checking the floor display or adjusting her bag, completely unaware of the camera.
The image should feel like a failed candid snapshot rather than a selfie.
LATE NIGHT GAS STATION
Create an ultra-realistic accidental front-camera photo at a late-night gas station.
The phone accidentally captures an attractive young woman while walking back to her car.
Low-angle perspective, uneven artificial lighting, slight blur from movement, imperfect framing, natural smartphone distortion, and visible digital noise.
Background includes fuel pumps, parked cars, bright signs, and dark surroundings.
The image should feel raw, unplanned, and strangely believable.
GYM ACCIDENTAL SNAPSHOT
Create an ultra-realistic accidental front-camera smartphone photo captured inside a real gym.
An attractive young woman is unintentionally photographed while working out or moving between equipment. The phone appears to have activated accidentally while being carried, placed near a bench, picked up from the floor, or slipping from the hand.
Use a low-angle front-camera perspective with awkward framing, slight tilt, imperfect crop, natural smartphone distortion, and unstable composition.
The subject wears modern gym clothing with a naturally flattering athletic silhouette. She is not posing. She appears caught mid-movement while adjusting weights, walking past equipment, sitting on a bench, reaching for a water bottle, stretching, or turning naturally.
Her eyes are not locked on the camera. Expression feels spontaneous, focused, distracted, and completely unaware of the capture.
Include realistic gym details: workout machines, dumbbells, mirrors, benches, rubber flooring, water bottles, towels, fluorescent lighting, and other people softly blurred in the background.
Add subtle motion blur, camera shake, front-camera softness, smartphone noise, indoor sensor grain, slight compression artifacts, imperfect exposure, and realistic low-light gym ambience.
The final image should feel like a real accidental photo discovered later in a phone gallery, not a fitness photoshoot, not a posed gym selfie, and not social media content.
SUPERMARKET ACCIDENTAL SNAPSHOT
Create an ultra-realistic accidental front-camera smartphone photo captured inside a modern supermarket.
An attractive young woman is unintentionally photographed while shopping. The phone appears to have activated accidentally while being carried in one hand, near a shopping basket, cart, or jacket pocket.
Use a low-angle front-camera perspective with awkward framing, slight tilt, imperfect crop, and natural smartphone distortion.
The subject wears casual stylish everyday clothing with a naturally flattering silhouette. She appears caught while reaching for products, pushing a cart, checking shelves, comparing items, or walking through an aisle.
She is not looking at the camera and appears completely unaware of the capture.
Include realistic supermarket details: product shelves, shopping carts, fluorescent lighting, price tags, checkout areas, customers in the background, and everyday store clutter.
Add subtle motion blur, realistic smartphone grain, sensor noise, front-camera softness, slight compression texture, natural exposure fluctuations, and imperfect focus.
The image should feel ordinary, random, believable, and accidentally captured, like a forgotten image found later in a phone gallery rather than a staged photograph.
NIGHTCLUB ACCIDENTAL SNAPSHOT
Create an ultra-realistic accidental front-camera smartphone photo captured inside a crowded nightclub.
An attractive young woman is unintentionally photographed while moving through the crowd. The phone appears to have activated accidentally while being carried, lowered, or moved through the club.
The camera angle should be low, awkward, slightly tilted, and unstable, as if the phone was hanging near waist level or slipping in the hand.
The subject wears fashionable nightlife clothing with a naturally flattering silhouette and confident presence, but she is not posing. She appears caught mid-motion while dancing, turning, walking through the crowd, laughing with friends, or reacting naturally.
Her eyes are not locked on the camera. Expression feels spontaneous, distracted, and completely unaware of the capture.
Include realistic nightclub details: blurred people, colored practical lights, LED glow, dim ambient lighting, movement in the crowd, reflections, and nightlife atmosphere.
Add realistic motion blur, camera shake, front-camera softness, smartphone noise, low-light sensor grain, slight compression artifacts, imperfect exposure, and off-center framing.
The final image should feel like a real accidental photo discovered later in a phone gallery, not a planned selfie, not a professional shoot, and not social media content.
LIVING ROOM ACCIDENT
Create an ultra-realistic accidental smartphone snapshot in a living room.
An attractive young woman is captured while standing up from a sofa or walking through the room.
The phone appears to have activated accidentally while being carried.
Awkward low-angle composition, partial crop, slight motion blur, imperfect focus, smartphone grain, natural exposure fluctuations, and realistic front-camera softness.
Visible sofa, television glow, lamp lighting, cushions, and normal household clutter.
The image should feel like a completely ordinary accidental photo found in a gallery rather than a planned image.
Transform each uploaded photograph into a minimalist premium handmade paper illustration, using it as the sole reference for identity, pose, proportions, clothing, accessories, colors, and recognizable details.
Create ONLY the illustration. No original photograph or photographic section.
FORMAT LOCK — NON-NEGOTIABLE
Strict vertical 3:4.
Single full-frame minimalist editorial artwork.
Subject small, centered, and restrained, occupying 10–20% of the canvas with generous negative space.
IDENTITY & REFERENCE LOCK
Preserve:
Facial identity and structure, Face shape and hairstyle, Skin tone, Body proportions, Clothing and accessories, Pose or gesture, Important objects, Dominant colors, Overall visual character
Keep the subject immediately recognizable. Do not alter identity, anatomy, hairstyle, clothing, or add unrelated elements.
BOTTOM HALF — MINIMAL HAND-DRAWN PAPER ILLUSTRATION
Create a small, centered, highly recognizable handmade illustration.
Simplify aggressively while preserving essential features.
Use delicate imperfect lines, minimal expressive linework, few acrylic-style flat color shapes, visible brush marks, irregular organic edges, subtle pigment variation, tactile paper fibers, and natural handmade imperfections.
It must look physically painted and drawn on textured paper, not digitally generated.
BACKGROUND
Minimal rough white, warm off-white, or pale natural paper with subtle fibers and grain.
Keep it overwhelmingly empty. Only a few minimal lines or shapes may suggest the original environment.
No detailed or decorative background.
COLOR SYSTEM
Extract colors from the photograph.
Use maximum four main colors.
Keep the palette restrained, sophisticated, harmonious, and photographically connected.
Use flat acrylic-style shapes with paper grain, dry-brush texture, and pigment variation.
COMPOSITION & NEGATIVE SPACE
Keep the illustration precisely centered at 10–20% scale.
Surround it with generous negative space.
Mood: quiet, poetic, refined, luxurious, editorial, thoughtful, elegant, minimal, premium.
TYPOGRAPHY — STRICTLY PROHIBITED
ZERO TEXT.
No words, letters, numbers, titles, captions, labels, dates, logos, signatures, watermarks, or readable symbols.
VISUAL LANGUAGE
Quiet, poetic, refined, minimal, innocent, relaxed, artistic, thoughtful, elegant, recognizable, premium.
OUTPUT STYLE
Premium contemporary editorial / modern art-book illustration.
Tactile handmade paper, hand-drawn artwork, sophisticated visual storytelling.
No digital vector aesthetic.
FINAL RESULT
ONLY one small, centered, highly recognizable handmade paper illustration derived from the uploaded photograph, surrounded by generous textured-paper negative space.
Maximum four colors. No photograph. No photographic section. No text.
NEGATIVE PROMPT
No photograph, photographic section, split-screen, upper half, collage, multiple images, second subject, text, typography, letters, words, numbers, logos, watermark, signature, detailed background, crowded composition, oversized subject, excessive detail, altered identity, distorted face or anatomy, invented objects, chaotic colors, rainbow palette, complex gradients, glossy CGI, 3D rendering, digital vector art, sterile digital artwork, decorative elements, or readable symbols.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 1,851 characters
Show the full prompt
Transform the uploaded photo into a premium fashion editorial where the outfit is recreated as handmade paper cutouts layered on top of a fully realistic photograph.
FORMAT LOCK
Vertical 4:5 composition. Preserve the original camera angle, framing, perspective, body pose, subject placement, and overall scene structure. Do not crop or alter the composition.
IDENTITY LOCK
Keep the original photograph completely realistic. Preserve the exact face, hairstyle, skin tone, facial expression, makeup, hands, arms, legs, body proportions, lighting, shadows, and background. Do not stylize or redraw the person.
PAPER CLOTHING EFFECT
Transform only the clothing, shoes, bag, and wearable accessories into handcrafted paper cutouts. Each item should look as if it was hand-drawn on paper, carefully cut out with scissors, and physically glued back onto the photograph.
PAPER STYLE
Visible white paper texture. Colored-pencil strokes. Hand-colored imperfections. Soft marker shading. Wax crayon details. Slightly uneven scissor-cut edges. Layered paper collage effect. Natural paper thickness. Small handmade imperfections. Premium craft aesthetic.
STYLE
Luxury fashion editorial. Handmade mixed-media collage. Contemporary magazine cover. Premium paper craft illustration. Modern artistic styling. High-end fashion campaign. Clean minimalist composition. Museum-quality mixed-media artwork.
OUTPUT
Ultra-detailed 8K quality. Photorealistic model with handcrafted paper clothing. Elegant editorial design. Premium visual storytelling.
NEGATIVE PROMPT
Avoid changing the face, hairstyle, skin, hands, body proportions, lighting, shadows, or background. Avoid digital filter effects, fully illustrated characters, CGI appearance, plastic textures, blurry details, messy cutouts, perfect vector edges, text, logos, watermarks, low quality, or distorted anatomy.
Prompt:
Use the uploaded photo as the identity reference for the main person.
Transform the person into a cinematic editorial photograph set in a crowded urban environment during warm golden hour.
IMPORTANT:
Preserve the person’s recognizable facial identity, natural facial proportions, skin tone, eyes, nose, lips, and overall likeness from the uploaded reference photo.
Place the person standing calmly in the center of a busy moving crowd. The main person must remain relatively sharp and visually clear while everyone around them is captured with strong horizontal motion blur, creating the feeling that the crowd is rushing past.
Keep the main person’s natural pose and body proportions believable. Adapt the clothing naturally to a simple elegant dark outfit or coat suitable for the scene.
Create:
- warm golden-hour sunlight
- soft cinematic backlighting
- subtle rim light around the hair
- shallow depth of field
- realistic skin texture
- natural hair detail
- atmospheric urban background
- heavily motion-blurred pedestrians surrounding the subject
- foreground silhouettes and blurred figures partially crossing the frame
- warm amber, brown, and muted dark tones
- professional cinematic photography
- realistic camera optics
- natural film-like texture
- emotional, quiet, introspective atmosphere
The central person should feel completely still and composed while the surrounding world appears to move rapidly around them.
The face must be naturally integrated into the scene with matching lighting, shadows, perspective, focus, and color grading.
Do not make the image look like an illustration, painting, cartoon, 3D render, or artificial AI artwork.
Do not add text, logos, watermarks, borders, frames, or extra focus on the background people.
The final result should look like a professionally photographed cinematic moment of the person standing still in a rushing crowd, matching the same overall visual concept and mood regardless of whether the uploaded reference is a man or woman.
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Visible excerpt, verbatim · Copy gives these lines only
Create a premium editorial sticker-collage portrait using the uploaded image as the only identity reference.
FORMAT LOCK
Vertical 4:5 composition, centered head-and-shoulders portrait, premium fashion-editorial framing, clean blue-sky background.
IDENTITY LOCK
Prompt:
A stylish mixed-media collage portrait of a young woman sitting cross-legged on the ground, smiling warmly with her chin resting on her hand. She has dark braided hair in twin French braids, wearing an oversized black crewneck sweatshirt, light grey cuffed jeans, and clean white sneakers. In the background, there is a large faded black-and-white double-exposure portrait of her smiling. The backdrop features artistic grunge textures, abstract black paint brushstrokes, distressed canvas, subtle vintage cursive handwriting, and bold brush-typography stencil text reading "HANIA" at the top left. Neutral color palette with warm beige tones, soft studio lighting, highly detailed photo edit, 4:5 aspect ratio --ar 4:5
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,485 characters
Show the full prompt
Transform the uploaded image into a premium handcrafted textile artwork while preserving the original composition, subject, pose, camera angle, framing, perspective, and overall scene structure.
FORMAT LOCK
Preserve the original composition exactly. Do not change the camera angle, crop, subject placement, proportions, background layout, or object positions. Maintain the original aspect ratio.
IDENTITY LOCK
If the image contains a person, use the uploaded photo as the only identity reference. Preserve the exact facial structure, facial proportions, hairstyle, expression, body proportions, clothing silhouette, and recognizable identity. Do not beautify, redesign, or reinterpret the subject.
TEXTILE TRANSFORMATION
Transform the entire image into a handcrafted embroidery artwork. Every object, person, and background element should appear entirely constructed from threads, yarn, embroidery stitches, punch-needle loops, fabric appliqué, and layered textiles.
TEXTURE
Dense embroidered stitches, raised thread loops, visible yarn fibers, decorative stitched outlines, tufted embroidery, layered fabric pieces, tactile woven surfaces, handmade imperfections, dimensional relief, premium artisan craftsmanship.
COLOR PALETTE
Rich yet soft colors. Slightly muted tones. Harmonious handcrafted palette inspired by premium embroidery, textile illustration, and decorative fiber art.
BACKGROUND
Transform the entire background into embroidered fabric while preserving the original environment, depth, layout, and perspective. Buildings, furniture, landscapes, plants, skies, and objects should all become stitched textile elements with layered embroidery details.
STYLE
Luxury embroidery illustration, punch needle artwork, tufted textile collage, handcrafted tapestry, premium fiber art, decorative embroidery, cozy artistic aesthetic, whimsical yet elegant, museum-quality textile design.
OUTPUT STYLE
Ultra-detailed handcrafted embroidery artwork. Visible thread texture. Soft volumetric stitches. Layered fabric construction. High-relief textile surface. Premium artisan quality. The final image should look like a real embroidered masterpiece created entirely from fabric and thread.
NEGATIVE PROMPT
Avoid photorealistic surfaces, plastic materials, smooth digital painting, CGI appearance, blurry stitches, flat textures, missing thread details, identity changes, distorted anatomy, low-quality embroidery, random objects, text, logos, watermarks, or unrealistic fabric effects.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,041 characters
Show the full prompt
Transform the uploaded image into a premium surreal fashion editorial while preserving the exact facial identity, facial features, hairstyle, skin tone, body proportions, outfit, accessories, pose, camera angle, composition, and overall appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Preserve the original framing, perspective, subject placement, and camera distance. Keep the subject as the clear focal point.
IDENTITY LOCK
Preserve the exact face, hairstyle, skin texture, facial expression, outfit, accessories, colors, pose, and body proportions. Do not redesign the person, replace clothing, or alter any facial details.
CHROME SCULPTURES
Surround the subject with elegant organic liquid chrome structures inspired by flowing molten metal, biomorphic forms, and futuristic sculptures. The metallic elements should naturally wrap around the subject without covering the face or important body parts. Create realistic reflections, soft shadows, depth, perspective, and believable interaction between the chrome and the environment.
BACKGROUND
Maintain a clean minimalist background while seamlessly integrating the chrome sculptures into the scene. Preserve the original atmosphere and allow generous negative space for a refined editorial composition.
LIGHTING & STYLE
Soft natural daylight. Cinematic contrast. Subtle metallic highlights. Luxury fashion magazine aesthetic inspired by high-end editorials. Premium color grading. HDR realism. Crisp textures. Dramatic depth of field. Professional commercial photography. Ultra-realistic 8K rendering.
OUTPUT
Modern surreal fashion campaign. Photorealistic subject combined with sculptural liquid chrome elements. Elegant, minimal, futuristic, and museum-quality editorial artwork.
NEGATIVE PROMPT
Avoid changing the face, hairstyle, outfit, accessories, colors, pose, camera angle, or composition. Avoid extra people, text, logos, watermarks, clutter, unrealistic reflections, plastic-looking chrome, low quality, blur, distorted anatomy, or CGI-looking skin.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 3,093 characters
Show the full prompt
Create an ultra-realistic high-fashion surreal angel portrait of the uploaded person seated directly on the actual floor inside {{Scene}}, with massive {{Wing Color}} angel wings and a matching {{Outfit Color}} editorial outfit.
FORMAT LOCK
Vertical 4:5 composition. Subject seated in the lower half of the frame. Subject centered. Massive wings dominate the upper frame. Natural 35–50mm perspective. Moderate depth of field. Premium cinematic fashion photography.
GROUND CONTACT LOCK
The subject must always sit directly on the actual floor of {{Scene}}. Never sit on chairs, benches, subway seats, laundry machines, furniture, platforms, or props. Never float, hover, stand, or kneel. Clothing and skirt must naturally spread across the floor with clear physical contact.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve exact facial structure, facial proportions, eyes, nose, lips, jawline, hairstyle, hair color, skin tone, age appearance, body proportions, and recognizable identity. Do not beautify, redesign, replace, age, de-age, or reinterpret the person.
POSE SYSTEM
The subject is seated on the floor with an elegant tired posture. One hand may rest gently on the cheek, knee, or floor. Head may face forward, turn slightly left, turn slightly right, or look downward. Expression should feel subtle, melancholic, calm, and editorial. Pose can vary naturally, but the body must remain seated on the floor.
WARDROBE
{{Outfit Style}}. High-fashion surreal editorial styling. Fabric should feel detailed, textured, layered, and premium. Outfit color must match {{Outfit Color}}. Clothing should fall naturally onto the floor.
ANGEL WINGS
Massive voluminous {{Wing Color}} angel wings extending behind the subject. Wings spread wide in a dramatic symmetrical V-shape, framing the body and dominating the upper composition. Ultra-realistic feather texture, cinematic scale, soft glow or shadow depending on the scene.
ENVIRONMENT
Place the subject inside {{Scene}}. Include {{Location Details}}. The environment should feel cinematic, realistic, atmospheric, and visually connected to the mood. The floor must be visible under the subject.
LIGHTING
{{Lighting}}. Strong contrast between the subject, wings, outfit, and environment. Use cinematic highlights, realistic shadows, and subtle reflections when appropriate.
MOOD
{{Mood}}. Elegant, dreamlike, introspective, surreal, and editorial. The image should blend angelic fantasy with real-world cinematic atmosphere.
OUTPUT STYLE
Ultra-realistic cinematic fashion photography. Luxury editorial campaign. Hyperrealistic textures. Dramatic angelic surrealism. Premium magazine-quality image. 4K / 8K detail.
NEGATIVE PROMPT
Avoid floating body, hovering pose, sitting on chairs, sitting on benches, sitting on subway seats, sitting on laundry machines, standing, kneeling, changed identity, different face, beautified face, cartoon style, CGI look, distorted anatomy, extra limbs, damaged wings, tiny wings, messy feather artifacts, flat lighting, crowded background, low quality, blur, text, watermark, logo.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 3,758 characters
Show the full prompt
Create a hyper-realistic monochrome dark fashion editorial portrait featuring the uploaded subject standing alone in a vast empty natural landscape, surrounded by numerous black crows flying through the frame.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 composition.
Premium editorial portrait framing.
Large negative space above the subject.
Subject centered with clean silhouette separation.
Do not generate landscape, square, cinematic widescreen, or cropped compositions.
IDENTITY LOCK — NON-NEGOTIABLE
Use the uploaded character reference as the only identity source.
Preserve the subject’s exact facial structure, facial proportions, skin tone, eye shape, nose shape, lips, hairstyle, hair color, age appearance, skin texture, body proportions, and all recognizable identity traits.
The subject must remain clearly identifiable as the same person.
Do not alter, beautify, redesign, stylize, age, or reinterpret the person’s identity.
SCENE DIRECTION
Place the subject alone in a cold, open natural landscape with a distant horizon.
The environment should feel empty, quiet, isolated, and mysterious.
Use a barren field, wild grassland, empty moor, or open countryside setting.
No buildings, roads, vehicles, fences, city elements, or distractions.
The scene should feel inspired by gothic fine-art photography and Scandinavian noir, but not copy any specific image directly.
WARDROBE
Dress the subject in a premium black wool overcoat layered over a black turtleneck.
Natural fabric texture.
Subtle wind movement in the coat and hair.
Luxury monochrome styling.
Minimal accessories.
No bright clothing or colorful details.
CROWS & MOTION
Surround the subject with numerous realistic black crows flying naturally through the scene.
Some crows should pass close to the camera.
Some should fly behind the subject.
Some should appear smaller in the distance.
Use varied wing positions and natural flock movement.
Add slight motion blur to selected foreground birds.
Keep the subject sharp, calm, and motionless while the birds create controlled visual chaos.
Crows must look realistic, not fantasy creatures.
EXPRESSION & POSE
The subject stands still with quiet confidence.
Neutral, emotionally detached expression.
No smile.
No social media pose.
No exaggerated fashion pose.
Hands may rest naturally in the coat pockets.
The subject may look slightly off-camera or into the distance.
LIGHTING & WEATHER
Soft overcast daylight.
Cold white-grey cloudy sky.
No visible sun.
No dramatic sunset.
No blue sky.
Diffused natural lighting with high-end editorial contrast.
Subtle rim separation around the subject.
CAMERA & COMPOSITION
Telephoto lens compression.
Fashion magazine framing.
Shallow depth of field.
Large sky area above the subject.
Balanced crow placement throughout the frame.
Minimalist composition with cinematic depth.
The subject should feel powerful, isolated, and still.
COLOR GRADING
Monochrome black-and-white grading.
Rich blacks.
Soft silver highlights.
Deep contrast.
Subtle film grain.
Fine-art photographic finish.
No colorful elements.
OUTPUT STYLE
Luxury dark fashion campaign.
Gothic fine-art portrait.
Scandinavian noir mood.
Hyper-realistic editorial photography.
Monochrome cinematic realism.
8K quality.
The final image should feel like a powerful luxury fashion campaign mixed with haunting black-and-white fine-art photography, not a fantasy poster or AI-generated scene.
NEGATIVE PROMPT
Avoid colorful elements, bright clothing, smiling expression, fantasy birds, cartoon style, CGI appearance, plastic skin, changed identity, distorted face, extra limbs, distorted anatomy, unrealistic feathers, cluttered background, buildings, roads, vehicles, fences, text, watermark, logo, low quality, oversharpening, and AI artifacts.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,610 characters
Show the full prompt
Transform the uploaded person into a premium cinematic sports fashion campaign while preserving the exact facial identity, facial structure, skin tone, hairstyle, eye shape, nose, lips, jawline, body proportions, pose, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Dynamic mid-air action pose. Premium sports campaign photography. Centered editorial composition.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial features, hairstyle, skin tone, facial proportions, body proportions, pose, and overall appearance. Do not redesign or alter the person’s identity.
ACTION POSE
Capture the subject suspended mid-air in a graceful athletic jump. The body is slightly rotated sideways with the torso leaning diagonally forward. The head tilts slightly downward with a calm, focused expression. One leg bends upward beneath the body while the other extends diagonally backward with a naturally pointed foot. One arm reaches downward with relaxed fingers while the opposite arm extends backward for balance. Hair flows naturally upward and backward, emphasizing realistic movement and momentum.
STYLING
Oversized premium white Nike sweatshirt. Semi-transparent black-and-white marble smoke jogger pants. Futuristic white Nike Air sneakers with glowing violet air soles. Keep realistic fabric textures, folds, and premium sportswear quality.
ENERGY EFFECTS
Wrap the subject with swirling luminous purple energy smoke around the legs, waist, arms, and hair. Blend the energy naturally into the movement with glowing violet particles, realistic volumetric smoke, subtle light trails, and dynamic motion.
BACKGROUND
Deep black-to-purple gradient background. Massive metallic chrome “JUST DO IT” typography behind the subject with a large silver Nike swoosh positioned above. Clean luxury sports campaign composition with realistic reflections and generous negative space.
LIGHTING & STYLE
IMAX-level cinematic lighting. Dramatic rim lighting. Volumetric light beams. High contrast. HDR rendering. Ultra-realistic skin texture. Sharp fabric detail. Premium commercial sports advertising aesthetic. Luxury editorial photography. Vibrant purple glow. Photorealistic 8K quality.
NEGATIVE PROMPT
Avoid standing poses, sitting poses, stiff body posture, incorrect limb positions, unrealistic floating, distorted anatomy, extra limbs, blurry details, weak energy effects, flat lighting, dull colors, cropped body, low quality, watermarks, text artifacts, logos outside the intended Nike branding, or unrealistic rendering.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,842 characters
Show the full prompt
Transform the uploaded person into a premium retro 1980s fashion editorial poster while preserving the exact facial identity, facial structure, hairstyle, skin tone, facial proportions, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Centered framing. Subject fully visible from head to shoes. Premium fashion campaign poster.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the exact facial features, hairstyle, skin tone, body proportions, and recognizable appearance. Do not redesign the person’s identity.
POSE & OUTFIT
Confident full-body pose with hands naturally placed inside the pockets of relaxed high-waisted white denim jeans and legs slightly apart. Dress the subject in an oversized vintage 1980s color-block windbreaker featuring lavender, purple, teal, and hot pink over a clean white crew-neck T-shirt. Add white high-top canvas sneakers and white socks. Keep realistic fabric textures and natural clothing folds.
BACKGROUND
Bold electric cobalt-blue seamless studio backdrop with subtle deep-blue tonal variation, creating a striking contrast against the lavender, purple, teal, and hot-pink color-block windbreaker. Use dramatic directional studio lighting with a strong cinematic shadow cast behind the subject. Add a subtle retro neon glow around the silhouette while keeping the background clean, polished, and visually controlled. The cobalt-blue backdrop should create a vivid 1980s synth-pop atmosphere without competing with the subject.
RETRO DOODLES
Surround the subject with hand-drawn white chalk-style 1980s doodles including cassette tapes, oversized headphones, a classic boombox, vinyl records, musical notes, stars, lightning bolts, arrows, zigzag lines, geometric shapes, and energetic scribbles. Keep the graphics balanced around the subject without covering the face or body.
TYPOGRAPHY
Add authentic handwritten retro typography reading “A MIX TAPE” in the upper-left and “80’s VIBES” on the right side. Keep the lettering sharp, readable, correctly spelled, and naturally integrated into the doodle artwork.
LIGHTING & STYLE
High-fashion streetwear campaign aesthetic. Nostalgic synth-pop energy. Vibrant pink and purple palette. Realistic skin and hair. Crisp clothing textures. Sharp focus. High contrast. Soft studio highlights. Mixed-media photography and hand-drawn illustration. Premium magazine advertising aesthetic. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid changing identity, facial features, hairstyle, skin tone, body proportions, blurry face, distorted anatomy, extra fingers, extra limbs, poorly drawn hands, warped clothing, duplicate objects, dull colors, cluttered composition, unreadable text, spelling errors, text artifacts, watermark, logo, cropped feet, or low-quality rendering.
7 examples are not shown here because they show real public figures. See all on the site ↗
Luxury black polo studio portrait
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,034 characters
Show the full prompt
Create an ultra-realistic luxury studio fashion portrait of the uploaded person wearing a plain black polo shirt.
FORMAT LOCK
Vertical 4:5 composition. Waist-up portrait. Subject centered. Premium commercial campaign framing. Magazine-cover quality.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the exact facial structure, facial proportions, hairstyle, hair color, skin tone, expression, body proportions, and all recognizable identity traits. The face must remain 100% recognizable.
WARDROBE
Plain black polo shirt. Clean premium fit. No logos. No graphics. Luxury commercial apparel styling.
POSE
The subject gazes slightly upward and away from the camera with a confident cinematic expression. Relaxed posture. Strong fashion-editorial presence.
STUDIO SETTING
Vibrant fiery orange-to-red seamless studio backdrop with a smooth radial gradient, brighter behind the subject and fading into deeper crimson-red tones toward the edges. Clean studio environment with no props or distractions.
LIGHTING
Intense golden rim lighting around the hair and shoulders. Dramatic backlight. Soft beauty key light illuminating the face. High-contrast commercial lighting. Warm cinematic color grading. Natural skin rendering. Soft shadow transitions.
CAMERA
Medium-format fashion photography. 85mm lens. f/2.0 aperture. Shallow depth of field. Razor-sharp facial focus. HDR rendering. Fine film grain.
OUTPUT STYLE
Luxury commercial advertising campaign. Premium fashion editorial photography. Ultra-realistic skin texture. Clean minimal composition. Magazine-quality finish. Powerful cinematic mood. Ultra-detailed 8K rendering.
NEGATIVE PROMPT
Avoid blurry details, low resolution, image noise, overexposed highlights, underexposed shadows, distorted face, deformed features, extra limbs, duplicate body parts, bad anatomy, unrealistic skin texture, cartoon style, painting, CGI, oversaturated colors, motion blur, soft focus, poor lighting, low detail, artifacts, text, logos, watermarks, cluttered background.
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,175 characters
Show the full prompt
PROMPT SETTINGS
Balloon Number: 19
Balloon Color: Metallic Silver, Chrome Silver, Matte Black, Pearl White, Champagne Gold, Royal Blue, Emerald Green, Orange, Blush Pink, Lavender Purple, Sunflower Yellow
________________
Upload one reference image before generating.
MAIN PROMPT
Transform the uploaded person into a premium luxury fashion editorial portrait while preserving the exact facial identity, facial structure, hairstyle, skin tone, body proportions, outfit style, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Clean white seamless studio background. Centered composition. High-end fashion campaign aesthetic.
IDENTITY LOCK
Preserve the exact facial features, hairstyle, skin tone, facial proportions, body proportions, expression, and overall identity from the uploaded image. Do not redesign the face or alter the person's appearance.
MAIN CHARACTER
The subject stands confidently against a clean studio backdrop, holding oversized [Balloon Color] metallic foil balloons shaped as "[Balloon Number]". The pose should feel relaxed, fashionable, and editorial with subtle asymmetry, natural body language, and confident energy.
OUTFIT
Preserve the outfit from the uploaded reference image. If no outfit is visible, dress the subject in a tailored black blazer, high-waisted tailored trousers, and elegant minimalist styling.
DETAILS
Realistic metallic balloon material with accurate reflections, natural folds, proper inflation, realistic string tension, subtle hair movement, natural hand placement, clean composition, and authentic balloon physics.
LIGHTING & STYLE
Soft professional studio lighting. Clean natural shadows. Editorial fashion photography. Premium magazine campaign aesthetic. Ultra-realistic skin texture. High-detail fabric rendering. Luxury color grading. Sharp focus. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid blurry details, distorted face, changed identity, altered outfit, unrealistic balloon shapes, incorrect reflections, extra fingers, extra limbs, bad anatomy, cluttered background, harsh lighting, plastic skin, text, logos, watermarks, or low-quality rendering.
Text from the author's doc; differs slightly from the site.
Fine-art fashion sketch portrait
Figure potential: medium (outfit/pose reference for a figure)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Illustration styles
Full prompt, verbatim · 2,454 characters
Show the full prompt
Transform the uploaded portrait into a premium hand-drawn fine-art fashion sketch while preserving the exact facial identity, facial structure, hairstyle, proportions, expression, and recognizable likeness.
FORMAT LOCK
Vertical 3:4 composition. Single centered portrait. Clean sketchbook layout. The artwork should fade naturally into the paper with no hard borders.
IDENTITY LOCK
Preserve the exact facial features, eye shape, nose, lips, jawline, hairstyle, skin tone, facial proportions, and expression from the uploaded image. Do not redesign or stylize the identity.
PAPER & MEDIUM
Aged warm off-white kraft paper with subtle visible paper fibers and natural grain. Use a strict monochrome palette with rich black, soft charcoal grays, and the paper color as the brightest highlight. No pure white or additional colors. Apply only a subtle soft dark forest-green accent to the clothing while keeping the rest of the artwork monochrome.
DRAWING STYLE
Traditional hand-drawn portrait combining graphite pencil, soft charcoal, and expressive black ink. Hair created with long flowing energetic strokes, loose flyaway strands, varied line weight, and spontaneous movement. Facial features rendered with delicate disappearing contour lines and refined realistic detail. Clothing illustrated with bold broken brush strokes, dry-brush texture, unfinished edges, and subtle construction sketch lines.
SHADING
Large shadow areas built with layered charcoal and diluted ink washes. Expressive hatching and cross-hatching follow the form naturally. Skin remains smoothly blended with soft tonal transitions. Strong contrast between deep blacks and gentle gray gradients while preserving the paper texture.
COMPLETION EFFECT
The face is the most refined and detailed area. Moving outward toward the hair, shoulders, and clothing, the drawing gradually becomes looser, rougher, and more abstract, with strokes naturally dissolving into the paper for an authentic sketchbook aesthetic.
STYLE
Luxury fashion illustration. Fine-art academic sketch. Handmade mixed-media drawing. Editorial fashion artwork. Museum-quality print aesthetic. Expressive, elegant, and ultra-detailed craftsmanship.
NEGATIVE PROMPT
Avoid digital painting, vector graphics, plastic textures, excessive smoothing, bright colors, watercolor effects, anime style, cartoon rendering, blurry details, distorted anatomy, altered facial identity, text, logos, watermarks, or hard artwork borders.
{
"prompt_type": "reference_template_identity_replacement",
"CORE_RULE": "Use the supplied image as a locked photographic template. Preserve the photograph as closely as possible, but replace the woman's facial identity with a completely different fictional adult woman.",
"PRESERVE": [
"exact pose and body position",
"head angle and gaze direction",
"camera angle, height, perspective and framing",
"subject scale and crop",
"background and environment",
"lighting and shadows",
"hairstyle, hair length and placement",
"ALL clothing exactly as shown in the reference",
"ALL accessories and jewelry exactly as shown in the reference",
"glasses, if present",
"clothing colors, materials, patterns, fit and silhouette",
"jewelry type, placement and appearance",
"overall smartphone photography aesthetic"
],
"IDENTITY_CHANGE": {
"instruction": "Create a completely different fictional adult woman. Do not preserve the reference woman's recognizable identity or facial structure.",
"change": [
"face shape",
"eyes and eye spacing",
"eyebrows",
"nose",
"cheekbones",
"jawline",
"chin",
"lips",
"overall facial proportions"
],
"rule": "The new woman must clearly look like a different person while keeping the same photograph composition."
},
"CLOTHING_AND_ACCESSORIES": {
"instruction": "Do not invent, redesign or recolor the wardrobe. Automatically copy the clothing and accessories visible in the supplied reference image.",
"preserve": [
"top or shirt",
"bottoms",
"dress or outerwear if present",
"shoes if visible",
"glasses if present",
"necklaces",
"earrings",
"rings",
"bracelets",
"watches",
"all other visible accessories"
],
"important": "Whatever the woman is wearing or carrying in the reference must remain visually consistent in the generated image."
},
"PHOTOGRAPHY": {
"style": "authentic candid smartphone photograph",
"realism": "extreme photorealism",
"skin": "natural pores and realistic imperfections",
"processing": "natural smartphone HDR",
"retouching": "minimal and believable"
},
"CLEANUP": [
"remove text",
"remove watermarks",
"remove logos",
"remove captions",
"remove UI elements",
"remove screen borders",
"remove black strips or borders"
],
"NEGATIVE_PROMPT": [
"same woman",
"same face",
"facial duplicate",
"recognizable identity",
"different pose",
"different camera angle",
"different framing",
"different background",
"different clothing",
"different accessories",
"invented jewelry",
"invented clothing",
"changed clothing color",
"missing accessories",
"missing glasses",
"CGI",
"3D render",
"plastic skin",
"airbrushed skin",
"anime",
"illustration",
"extra limbs",
"deformed hands",
"warped anatomy",
"text",
"watermark",
"logo",
"UI"
],
"FINAL_COMMAND": "Treat the supplied image as a locked photographic template, not an identity reference. Keep the exact pose, camera, composition, environment, lighting, hairstyle, clothing and every visible accessory from the reference. Replace only the woman's facial identity with a completely different fictional adult woman. The result must look like a genuine photograph of a different person wearing the exact same outfit and accessories in the exact same scene."
}
Visible excerpt, verbatim · Copy gives these lines only
{
"prompt_type": "photorealistic_reference_based_image",
"reference": "Use the uploaded image only as a visual reference for pose, body positioning, camera angle, composition, accessories and overall photographic style.",
"subject": {
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 2,405 characters
Show the full prompt
Create a hyperrealistic cinematic studio portrait using the uploaded image as the only identity reference. Preserve the exact facial identity, facial structure, hairstyle, skin tone, proportions, and recognizable features.
FORMAT LOCK
Vertical 3:4 composition. Chest-up portrait. Subject perfectly centered and symmetrical. Head facing directly toward the camera, chin level, both shoulders fully visible.
IDENTITY LOCK
Preserve the exact face, eye shape, nose, lips, jawline, hairstyle, natural hair volume, skin texture, and facial proportions. Do not beautify, redesign, age, or alter the identity.
EXPRESSION & CLOTHING
Calm, neutral, slightly serious expression. Mouth closed, relaxed lips, direct attentive eyes. Plain pure-white round-neck T-shirt or oversized hoodie, loose relaxed fit, no logos or prints. The fabric should have a subtle realistic sheen under the studio lighting.
LIGHTING
Double-tone cinematic lighting. Warm amber-orange rim light from camera-left grazing one side of the face, neck, hair, and shoulder. Cool blue ambient light illuminating the rest of the face and body. Strong warm-versus-cool contrast with soft studio diffusion, clean highlights, and gentle shadows without obscuring facial features.
BACKGROUND
Seamless studio backdrop with a smooth deep-navy-to-teal gradient, darker toward the edges with subtle corner vignetting. Soft blue atmospheric glow. No props, patterns, visible texture, or distractions.
CAMERA & DETAIL
85mm portrait-lens aesthetic. Shallow depth of field similar to f/2.8. Eyes and facial features perfectly sharp. Medium-format digital photography look. Natural pores, subtle skin sheen, realistic imperfections, individual hair strands, crisp clothing texture, cinematic teal-and-orange color grading, filmic contrast, professional retouching while preserving natural skin.
OUTPUT STYLE
Indistinguishable from a real high-end studio photograph. Ultra-realistic commercial portrait. Premium editorial photography. HDR detail. Cinematic color grade. Photorealistic 8K quality.
NEGATIVE PROMPT
Avoid altered identity, different facial features, beauty filters, plastic skin, excessive skin smoothing, caricature, illustration, anime, CGI appearance, blurry eyes, distorted anatomy, asymmetrical framing, cropped shoulders, harsh facial shadows, logos, text, watermarks, extra people, props, reference-image overlays, or artificial-looking skin.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,627 characters
Show the full prompt
Create an ultra-realistic lifestyle editorial studio photo using the uploaded image as the only identity reference.
FORMAT LOCK
Vertical 4:5 composition. Clean minimalist white seamless studio background. Full-body composition. Main character centered. Premium editorial photography.
IDENTITY LOCK
Preserve the exact facial features, facial structure, hairstyle, skin tone, body proportions, outfit, shoes, colors, and overall appearance from the uploaded image. Do not redesign the character or change any clothing.
MAIN CHARACTER
Capture the subject mid-run in a natural, energetic, and believable running pose. The body leans forward with one foot lifted while both arms move naturally with the running motion. The upper body is slightly pulled backward as if the clothing is being tugged, while the legs continue moving forward. The expression should feel joyful, lively, and slightly panicked, laughing while trying to escape.
SEMI-CHIBI CHARACTER
Place a stylized 3D semi-chibi version of the same person on the left side. Preserve the recognizable identity while giving the character a larger head, smaller body, expressive eyes, and playful Pixar-inspired proportions. The mini character looks frustrated and determined, grabbing the main character's clothing with both hands while pulling backward with visible effort.
FABRIC PHYSICS
Create realistic fabric tension where the clothing is being pulled. Show stretched material, natural folds, wrinkles, deformation, and believable cloth simulation. The pulling force should visibly affect the upper body while the running momentum continues forward.
INTERACTION
The realistic person and the stylized semi-chibi version must interact naturally with convincing physics, dynamic movement, clear storytelling, and humorous energy. The main character remains the focal point while the chibi character enhances the action.
LIGHTING & STYLE
Soft professional studio lighting. Clean natural shadows. Sharp focus. Realistic skin texture. High-detail fabric. Cinematic editorial photography. Premium realism. Dynamic motion. Luxury magazine campaign aesthetic. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid low quality, blur, pixelation, noise, bad anatomy, extra fingers, missing fingers, distorted hands, broken joints, stiff poses, weak motion, incorrect fabric physics, flat clothing, missing wrinkles, unrealistic interaction, changed identity, changed outfit, changed shoes, extra characters, cluttered background, colored background, studio props, harsh shadows, dark lighting, plastic skin, poor 3D rendering, distorted face, empty expression, text, logos, watermarks.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Fashion editorials
Full prompt, verbatim · 2,181 characters
Show the full prompt
Transform the uploaded person into a premium cinematic editorial scene while preserving the exact facial identity, facial structure, hairstyle, skin tone, facial proportions, expression, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Centered composition. Cinematic storytelling. Luxury fashion editorial photography.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial features, hairstyle, skin tone, facial proportions, and overall appearance. Keep the gold hoop earrings exactly as shown in the reference image.
SCENE
Create an ultra-realistic scene in a vast desert during golden hour. The subject sits calmly on a small red folding chair while reading a newspaper with a peaceful, emotionless expression. The posture should feel relaxed and effortless, creating a striking contrast with the dramatic scene behind.
WARDROBE
Premium black knitted sweater, beige wide-leg trousers, gold hoop earrings, and clean white sneakers. Preserve realistic fabric textures and elegant styling.
BACKGROUND
Behind the subject, place an old vintage light-blue sedan completely engulfed in massive flames. Thick black smoke rises dramatically into the sky with realistic fire, glowing embers, heat distortion, drifting ash, and dust blending naturally into the desert landscape. The burning car should remain behind the subject without distracting from the main composition.
LIGHTING & STYLE
Warm golden-hour cinematic lighting. Orange fire glow. Volumetric smoke. HDR rendering. Natural color grading. Ultra-realistic fire and smoke simulation. Realistic skin texture. Shallow depth of field. Luxury editorial photography. Award-winning cinematic composition. Premium commercial quality. Ultra-detailed 8K rendering.
NEGATIVE PROMPT
Avoid changing the face, hairstyle, skin tone, body proportions, or identity. Avoid extra people, extra limbs, extra fingers, distorted anatomy, blurry details, cropped body, unrealistic fire or smoke, duplicate objects, oversaturated colors, cartoon style, anime, CGI appearance, text, logos, watermarks, noise, low resolution, or low-quality rendering.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 1,919 characters
Show the full prompt
Transform the uploaded person into a premium monochrome fine-art portrait while preserving the exact facial identity, facial structure, hairstyle, skin tone, facial proportions, expression, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Centered portrait. Clean white background. Elegant fine-art composition with generous negative space.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial features, eye shape, nose, lips, jawline, hairstyle, expression, and overall appearance. Do not redesign or alter the person’s identity.
DRAWING STYLE
Create an ultra-realistic black-and-white artistic portrait using refined charcoal and graphite techniques. Combine realistic facial anatomy with soft tonal transitions, elegant pencil hatching, delicate cross-hatching, subtle graphite textures, and ultra-fine hand-drawn linework. Preserve realistic skin texture, detailed hair strands, and natural facial depth.
SMOKE DISSOLUTION
Allow the portrait to gradually dissolve into flowing smoke, mist, fine particles, and ethereal textures. Hair strands should naturally blend into the surrounding atmosphere, creating a seamless transition between realism and abstraction. Keep the effect soft, elegant, airy, and physically believable without obscuring the face.
LIGHTING & STYLE
Soft ambient studio lighting. High-contrast monochrome tones. Smooth gray gradients. Museum-quality fine-art illustration. Timeless minimalist aesthetic. Calm, poetic, and introspective mood. Hyperrealistic portrait blended with expressive smoke-dissipation effects. Luxury editorial artwork. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid color, harsh outlines, digital painting, CGI appearance, plastic skin, exaggerated smoke, distorted anatomy, extra facial features, blurry details, cluttered composition, text, logos, watermarks, or low-quality rendering.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 2,403 characters
Show the full prompt
Transform the uploaded person into a premium fine-art children’s studio portrait while preserving the exact facial identity, facial structure, skin tone, hairstyle, eye shape, nose, lips, jawline, expression, and recognizable appearance from the reference image.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Centered composition. Premium luxury editorial studio photography. Clean minimalist framing.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial features, facial structure, hairstyle, skin tone, facial proportions, expression, and overall appearance. Do not redesign or alter the person’s identity.
POSE & OUTFIT
The subject sits naturally on a seamless dark charcoal studio floor, resting one elbow on a raised knee while supporting the chin with one hand in a relaxed, confident pose. Wear an oversized light pastel pink button-up shirt with rolled sleeves, loose light-wash blue jeans, white ankle socks, chunky white-and-gray fashion sneakers, and decorate the hair with several tiny pastel pink flower clips.
BACKGROUND PORTRAIT
Create a large monochrome black-and-white side-profile portrait of the same person softly faded into the dark background. Position it in the upper-right portion of the composition, creating a refined double-exposure inspired fine-art effect while remaining clearly separated from the foreground subject. The foreground remains fully vibrant while the background portrait appears grayscale with reduced opacity.
BACKGROUND
Smooth dark gray-to-black seamless studio backdrop with subtle gradient lighting. Clean, elegant, and distraction-free.
LIGHTING & STYLE
Soft cinematic studio lighting using a large octabox from camera left with gentle fill lighting. Natural catchlights. Soft flattering shadows. Neutral luxury editorial color grading. Natural skin texture. Soft skin retouching. Rich contrast. Matte blacks. High dynamic range. Shallow depth of field. Premium commercial portrait photography. Ultra-photorealistic 8K quality. Magazine-cover aesthetic. Award-winning fine-art portrait.
NEGATIVE PROMPT
Avoid blurry details, low resolution, oversaturated colors, extra limbs, distorted anatomy, unrealistic facial features, crossed eyes, cropped feet, harsh shadows, noisy images, plastic skin, cartoon style, CGI, painting effects, text, logos, watermarks, background clutter, or low-quality rendering.
Visible excerpt, verbatim · Copy gives these lines only
Create a cinematic scene of a woman walking through a misty pine forest at sunrise. She is wearing a dark green coat, carrying a vintage leather notebook. Soft golden light, realistic photography, emo…
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Visible excerpt, verbatim · Copy gives these lines only
Show the full prompt
Use the uploaded character face reference as the hero subject, preserving facial structure, eyes, nose, lips, skin tone, age, and overall identity exactly as they are with zero distortion, beautificat…
Captured on a Leica SL3 with a 50mm Summilux lens at f/2.8 using authentic slow shutter crowd photography. Chest-level frontal composition with the subject perfectly sharp while commuters blur natural…
A crowded underground metro platform at night filled with rushing commuters moving in every direction while the uploaded character stands completely motionless in the center frame. The emotional tone…
Muted A24-inspired color grading with charcoal blacks, dirty olive undertones, faded tungsten highlights, desaturated concrete greys, subtle amber skin tones, soft fluorescent metro lighting, low satu…
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 2,584 characters
Show the full prompt
Create an ultra-realistic vanity mirror selfie inside a car using the uploaded image as the only identity reference.
FORMAT LOCK
Vertical 4:5 composition. Authentic smartphone selfie aesthetic. Natural social-media photograph. Realistic iPhone-camera quality.
IDENTITY LOCK
Preserve the exact facial structure, facial proportions, smile style, skin tone, hairstyle, hair texture, hair length, eyebrows, eyes, nose, lips, jawline, and overall identity. Do not beautify, redesign, or modify facial features. The final result must look like the exact same person photographed naturally.
COMPOSITION — CRITICAL
The selfie is captured through a car sun-visor vanity mirror.
The mirror occupies only the central portion of the frame.
The {{Car Interior Color}} car interior naturally fills the upper and lower areas of the image.
Framing should feel spontaneous, slightly imperfect, casual, and naturally captured rather than professionally staged.
SELFIE DETAILS
Subject holding a {{Phone Color}} smartphone very close to the mirror.
Phone partially covers the face naturally.
{{Sunglasses Color}} rectangular sunglasses partially cover one eye.
Subtle closed-mouth smile.
Relaxed and confident expression.
Tiny earrings visible.
Glossy nude nails visible.
Loose hair strands around the sunglasses.
Natural skin texture and pores preserved.
LIGHTING
{{Lighting}} sunlight entering softly through the car windows.
Warm highlights across skin, hair, sunglasses, and phone edges.
Gentle realistic shadows.
Natural light falloff.
Authentic smartphone-camera exposure.
COLOR GRADING
Muted warm tones.
Creamy color palette.
Softly desaturated colors.
Lifted blacks.
Natural skin tones.
Premium lifestyle-photography aesthetic.
CAMERA QUALITY
Modern iPhone-camera realism.
Slight natural softness.
Realistic reflections in the mirror.
Subtle sensor noise.
Authentic smartphone dynamic range.
Natural focus falloff.
No artificial sharpening.
MOOD
{{Mood}}
Effortless aesthetic.
Cozy.
Premium.
Candid.
Instagram-worthy.
Believable real-life snapshot.
OUTPUT STYLE
Ultra-realistic iPhone selfie, vanity-mirror photography, golden-hour car portrait, authentic social-media image, realistic smartphone photography, premium lifestyle aesthetic, highly detailed natural rendering.
NEGATIVE PROMPT
Avoid beauty filters, excessive retouching, HDR effects, cinematic grading, studio lighting, AI-generated appearance, unrealistic skin, exaggerated makeup, perfect symmetry, fashion-editorial styling, over-sharpening, distorted reflections, text, watermarks, or altered identity.
Make the aspect ratio 3:4
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Video game worlds
Full prompt, verbatim · 3,122 characters
Show the full prompt
Transform the uploaded person into a premium original black-and-crimson futuristic arachnid-inspired character in a cinematic gaming-room environment while preserving the person’s natural appearance and recognizable identity.
FORMAT LOCK
Vertical 3:4 composition. Full-body portrait. Subject completely visible from head to shoes. High-angle cinematic camera looking downward. Centered framing. Premium editorial gaming-room photography.
IDENTITY LOCK
Use the uploaded image as the only identity reference. Preserve the exact facial structure, hairstyle, skin tone, age, body proportions, expression, and overall recognizable appearance. Do not redesign, exaggerate, beautify, or alter the person’s natural features.
CHARACTER DESIGN
Create a completely original futuristic tactical outfit using matte black technical fabric, deep crimson accents, subtle geometric patterns inspired by arachnid structures, realistic stitching, layered materials, and understated protective panels. Add a small original geometric chest insignia with no recognizable superhero symbol. The design should feel like premium futuristic fashion-tech, completely distinct from existing movie, comic, or game costumes.
ENVIRONMENT
Premium modern gaming bedroom with charcoal interiors, crimson LED lighting, gaming desk, RGB keyboard and mouse, large monitor displaying abstract red geometric artwork, gaming chair, minimalist bed, collectible shelves, warm practical lighting, and dark wooden flooring. Keep the environment sophisticated, clean, cinematic, and uncluttered.
POSE
Subject stands naturally in the center. One hand rests naturally inside the outfit pocket while the other holds a gaming controller. Direct eye contact with the camera. Calm, confident expression. Natural posture, relaxed shoulders, believable anatomy, and realistic hand positioning.
CAMERA
High-angle cinematic perspective looking downward. Full-body framing with enough surrounding environment to establish the gaming room. Natural lens perspective. Realistic depth of field. Subject remains the primary focal point.
LIGHTING & STYLE
Soft cinematic key lighting. Deep controlled shadows. Subtle crimson ambient illumination. Natural skin tones. Gentle rim lighting. Realistic fabric reflections. Soft volumetric atmosphere. Premium editorial color grading. Photorealistic fashion photography. Extremely detailed textures. Sharp facial detail. Realistic materials. High-end cinematic gaming campaign aesthetic. Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid recognizable existing superhero costumes, movie costumes, comic-book designs, copyrighted character appearances, recognizable logos or symbols, altered facial structure, artificial beauty filters, exaggerated anatomy, extra fingers, extra limbs, distorted hands, warped controller, plastic skin, cartoon style, anime style, illustration style, excessive red lighting, clutter, text, watermark, cropped head, cropped feet, duplicate objects, or unrealistic shadows.
IMPORTANT
Create a completely original futuristic character design. Do not reproduce or imitate any existing superhero character or costume.
Posted on YouMind under the same author. It continues the site's first line word for word, with YouMind's default values filled in (wooden platform, blush-pink dress), but it does not have the site's second line, "Negative Prompt: ...".
Show the variant
Ultra-realistic IMAX-level Netflix-style cinematic surreal fashion portrait, 9:16 vertical composition, a dreamy high-end editorial scene featuring a beautiful young woman standing gracefully on a small floating wooden platform surrounded by soft layered clouds in the sky, while a massive semi-transparent close-up portrait of the same woman appears beautifully in the background clouds as an ethereal double-exposure style composition, the woman wearing elegant round black sunglasses and a flowing blush-pink satin slip dress with a smooth luxurious fabric drape, long soft wavy light-brown hair styled in polished loose waves with natural volume and gentle movement, her face showing calm confidence, soft feminine elegance and a graceful self-assured mood, with relaxed brows, softly defined lips and a subtle attractive expression, the smaller full-body figure standing in a poised fashion pose with one hand on her waist and the other relaxed naturally, the giant portrait behind her turned slightly sideways for a premium editorial feel, background filled with voluminous clouds and atmospheric sky depth, enhanced with dramatic high-contrast evening golden-hour lighting, rich sunset glow, warm amber highlights, glowing golden edges, soft orange-peach sky illumination and deeper creamy shadow contrast, strong cinematic light separation between highlights and shadows, luminous golden rim light on the hair and dress, soft mist around the floating platform, ultra-rich color grading with warm champagne tones, creamy beige clouds, golden highlights and elegant sunset contrast, photorealistic rendering, premium luxury fashion poster aesthetic, HDR, masterpiece quality, ultra-detailed, cinematic beauty atmosphere, sharp facial detail, dreamy yet powerful visual storytelling.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 2,704 characters
Show the full prompt
Transform the uploaded image into an ultra-realistic thermal infrared surveillance scan while preserving the original composition, subject structure, framing, camera angle, silhouettes, and overall scene geometry.
FORMAT LOCK — NON-NEGOTIABLE
Preserve the original image aspect ratio.
Preserve the original framing and composition.
Do not crop, zoom, reposition, or redesign the scene.
Do not change the subject placement or scene layout.
STRUCTURE PRESERVATION
Keep the main subject, background shapes, object placement, body pose, proportions, and environmental structure recognizable from the original image.
Preserve silhouettes and geometry.
Remove natural photographic colors and convert the entire scene into thermal imaging.
THERMAL COLOR SYSTEM
Use authentic false-color FLIR infrared mapping.
Cold areas should appear as deep navy, cobalt blue, dark blue, and cyan.
Warm areas should appear as yellow, orange, red, and white-hot highlights.
Heat distribution must feel realistic and physically believable.
Do not use random rainbow coloring.
Do not apply artistic color grading.
HEAT SIGNATURES
Create believable thermal gradients across the scene.
Add brighter heat signatures on naturally warm regions.
Add soft thermal bloom and infrared glow bleed around warm areas.
Use realistic heat diffusion instead of sharp artificial color patches.
Maintain clear subject readability through thermal contrast.
SURVEILLANCE CAMERA REALISM
Add subtle sensor noise.
Add scan grain.
Add digital interference.
Add faint compression artifacts.
Add slight low-light surveillance softness.
The image should feel captured by an advanced thermal imaging system, not digitally painted.
HUD INTERFACE
Overlay minimal tactical surveillance HUD elements.
Include scanner corner marks.
Add a subtle targeting reticle.
Add detection boxes around major heat signatures.
Add small telemetry-style text.
Add temperature-style indicators.
Keep the interface clean, technical, and minimal.
Do not cover important facial or subject details.
Avoid excessive UI clutter.
LIGHTING & CONTRAST
Use high-tech machine-generated thermal contrast.
No natural daylight colors.
No cinematic color grading.
No realistic RGB photography tones.
The image should feel analytical, tactical, and sensor-based.
OUTPUT STYLE
Ultra-realistic thermal infrared scan.
Authentic FLIR imaging aesthetic.
Classified surveillance footage look.
Drone thermal vision.
Cyber surveillance scan.
Experimental tech interface.
Military-style tactical imaging realism.
High-detail false-color infrared capture.
The final image should resemble advanced classified thermal surveillance footage while remaining faithful to the original image structure.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 2,659 characters
Show the full prompt
Transform the uploaded image into an ultra-realistic thermal infrared surveillance scan while preserving the exact original composition, framing, subject placement, camera angle, perspective, proportions, and scene structure.
FORMAT LOCK
Preserve the original aspect ratio, composition, crop, camera distance, perspective, subject scale, and environmental layout. Do not reposition, redesign, remove, or add major elements.
THERMAL IMAGING SYSTEM
Convert the entire image into authentic FLIR-style thermal imaging.
Cold regions should appear in deep navy, cobalt blue, cyan, and cool teal tones.
Warm regions should appear in yellow, orange, and red tones.
Hotspots should glow in bright orange and intense red.
Maintain realistic thermal gradients and believable heat distribution.
Avoid random rainbow coloration.
HEAT SIGNATURES
Generate realistic infrared heat behavior.
Soft thermal bloom around warm regions.
Natural heat diffusion.
Infrared glow bleed.
Temperature transitions should feel physically believable and sensor-generated.
Preserve recognizable silhouettes while removing natural photographic colors.
SURVEILLANCE INTERFACE
Add subtle military-style thermal HUD elements.
Scanner corner markers.
Targeting reticle.
Detection boxes.
Telemetry indicators.
Tracking graphics.
Temperature-style readouts.
Minimal technical overlays.
Keep interface elements clean and realistic.
Do not overcrowd the image.
SENSOR QUALITY
Subtle thermal camera grain.
Infrared sensor noise.
Digital interference artifacts.
Light signal distortion.
Authentic thermal-camera texture.
The image should feel captured by a real surveillance system rather than digitally color graded.
LIGHTING & CONTRAST
Machine-generated contrast.
High-tech imaging aesthetic.
Strong subject separation.
Realistic thermal intensity mapping.
Advanced FLIR sensor appearance.
STYLE
Military FLIR imaging.
Drone thermal vision.
Tactical surveillance footage.
Classified infrared scan.
Experimental defense technology aesthetic.
Cyber surveillance interface.
Advanced sensor visualization.
OUTPUT STYLE
Authentic thermal-camera screenshot.
Ultra-realistic infrared imaging.
Military-grade surveillance scan.
High-tech tactical visualization.
FLIR thermal footage aesthetic.
Professional sensor-system realism.
Ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid natural colors, rainbow gradients, artistic color grading, cartoon effects, sci-fi holograms, unrealistic heat maps, fantasy interfaces, cinematic lighting, photorealistic color photography, excessive HUD clutter, watermarks, logos, text artifacts, or distorted geometry.
Make the aspect ratio 3:4
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 3,055 characters
Show the full prompt
Transform the uploaded photograph into a pure LiDAR point-cloud visualization. Abandon the photographic appearance entirely. Reconstruct the entire scene as thousands of separate glowing laser-return points suspended in black space.
POINT-CLOUD STRUCTURE
Nothing continuous survives: no photographic surfaces, no skin, no fabric, no smooth materials, no solid forms, and no filled areas. Every visible element must be constructed exclusively from individual points with clearly visible black gaps between them. The scene should feel hollow and spatially transparent rather than solid.
DEPTH & DENSITY
Use point density according to LiDAR surface orientation and scanner visibility. Points are dense where surfaces directly face the scanner and progressively thin out toward grazing angles, edges, recesses, and occluded regions. Preserve genuine three-dimensional depth so the far side of the subject can remain faintly visible through the sparse near-side point shell.
SCANNER SIGNATURE
Make the LiDAR scanning origin visually apparent through perfectly clean radial sweep lines extending from the scanner position. Create hard, sharply defined empty shadow regions behind objects wherever the laser beam cannot reach. These occlusion shadows should reinforce the physical scanning geometry.
COLOR MAPPING
Color every point by distance using a cold scientific depth ramp: violet through cyan to green. Keep the color assignment consistent with spatial distance, producing a precise scientific LiDAR visualization rather than decorative coloring.
COMPOSITION & REFERENCE LOCK
Preserve the original subject’s shape, proportions, framing, camera perspective, spatial relationships, and overall composition exactly as they appear in the uploaded photograph. Transform only the visual representation into LiDAR data.
BACKGROUND
Pure deep black space surrounding the point cloud. No photographic background, environmental surfaces, gradients, textures, or atmospheric scenery. The black negative space must remain clearly visible between individual points.
VISUAL CHARACTER
Extremely clean, precise, high-density scientific LiDAR scan aesthetic. Individual laser returns must remain visibly separated. Strong depth perception, transparent point-shell construction, sparse peripheral returns, realistic occlusion behavior, and crisp radial scanning geometry.
FINAL RULE
The result must look like an authentic three-dimensional LiDAR acquisition of the original scene—not a photograph covered with dots. Every object must exist only as discrete spatial points and scan geometry, with nothing solid or continuous remaining.
NEGATIVE CONSTRAINTS
No photographic surfaces, no skin texture, no fabric texture, no solid objects, no continuous meshes, no smooth shading, no filled surfaces, no realistic photograph, no painterly rendering, no background scenery, no fog, no bloom-heavy effects, no random particles, no decorative stars, no excessive density, no loss of black gaps, no altered composition, no distorted proportions, no extra objects.
3:4 aspects ratio
Figure potential: high (3D character with clean sculptable forms)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 2,414 characters
Show the full prompt
Transform the uploaded image into a premium high-contrast monochrome particle-render artwork while preserving the original subject, pose, proportions, and overall composition.
FORMAT LOCK
Vertical 3:4 composition. Poster layout. Strong cinematic negative space. Subject remains clearly recognizable and positioned according to the reference.
PARTICLE STYLE
Reconstruct the entire subject from thousands of tiny glowing white particles and voxel-like points on a pure black background. Use dense particle clusters, point-cloud dispersion, halftone density gradients, scattered particles, and subtle noise trails to define the silhouette. Simplify fine details into clean geometric forms while keeping the subject recognizable.
VISUAL EFFECTS
Pure black background. White particle clusters forming the subject. Subtle floating debris and dissolving particles around the edges. Very sparse orange-red accent particles. Soft bloom around dense particle areas. Create depth through particle density, particle size, and controlled dispersion. Slight voxel and point-cloud appearance.
COLOR
Strict monochrome palette: black and white only, with extremely minimal orange-red accent particles. No other colors. Keep the overall image high contrast, minimal, and sophisticated.
COMPOSITION
Preserve the original subject’s general pose and composition while translating everything into particle geometry. Use generous negative space and a clean brutalist futuristic poster aesthetic. Keep the particle silhouette sharp enough to remain recognizable.
TYPOGRAPHY
Add minimal thin monospaced typography along the left and right edges. White text only, small and secondary to the artwork. Use abstract technical-style labels and numbers rather than large readable headlines.
STYLE
Futuristic AI architecture visualization. Experimental generative art. Point-cloud simulation. Procedural particle system. Volumetric dispersion. Brutalist graphic design. Cinematic minimalism. Abstract but recognizable. Ultra-detailed particle rendering. Premium gallery-poster aesthetic. 8K quality.
NEGATIVE PROMPT
Avoid realistic textures, painterly effects, colorful backgrounds, low contrast, photorealism, excessive orange particles, messy composition, extra objects, excessive typography, blurry particles, noisy artifacts, smooth solid rendering, gradients unrelated to particle density, distorted anatomy, changed pose, or low resolution.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 2,269 characters
Show the full prompt
Transform the uploaded person into a cinematic augmented-reality music experience while preserving the exact facial identity, hairstyle, facial structure, expression, body proportions, pose, clothing, camera angle, composition, and overall scene layout.
FORMAT LOCK
Preserve the original aspect ratio, framing, perspective, subject placement, camera distance, and composition. Keep the person as the central focus.
IDENTITY LOCK
Use the uploaded image as the only identity source. Preserve the exact facial structure, face shape, hairstyle, hair color, eyebrows, eyes, nose, lips, jawline, skin tone, age appearance, and recognizable identity traits.
AR MUSIC SYSTEM
Surround the subject with floating premium music-player interface cards inspired by modern streaming platforms. The cards orbit naturally around the subject at varying depths. Some cards appear in the foreground partially overlapping the subject, while others float behind and beside the subject.
INTERFACE DESIGN
Translucent frosted-glass panels, rounded corners, soft glowing edges, modern glassmorphism aesthetic, subtle reflections, premium UI design, clean typography, luxury digital-product styling.
SONG CARDS
Create floating music-player cards featuring:
{{Song 1}}
{{Song 2}}
{{Song 3}}
{{Song 4}}
{{Song 5}}
{{Song 6}}
{{Song 7}}
Include elegant album-art placeholders, playback controls, progress bars, waveform elements, and premium streaming-player styling.
LIGHTING
Natural realistic lighting, soft ambient glow, subtle reflections, clean premium atmosphere, no dramatic lighting.
DEPTH & MOTION
Strong depth of field, foreground cards slightly out of focus, background cards softly blurred, subtle floating motion, realistic AR-world depth.
STYLE
Cinematic augmented reality, premium music visualization, luxury streaming-platform aesthetic, modern glassmorphism design, Instagram-worthy creative portrait.
OUTPUT STYLE
Ultra-realistic AR portrait, floating music ecosystem, premium streaming-player interfaces, luxury social-media artwork, ultra-detailed 8K quality.
NEGATIVE PROMPT
Avoid cyberpunk lighting, dramatic shadows, neon overload, anime style, cartoon rendering, distorted face, duplicated cards, cluttered composition, watermark, text artifacts, logos, or altered identity.
6 examples are not shown here because they show real public figures. See all on the site ↗
Football TV broadcast fan screenshot
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Accidental snapshots
Full prompt, verbatim · 4,113 characters
Show the full prompt
Use the character reference as the fixed identity source and the jersey reference as the fixed wardrobe source.
MAIN PROMPT
Create an ultra-realistic live football TV broadcast screenshot featuring the referenced person sitting naturally among fans inside a packed stadium during a high-profile match.
IDENTITY LOCK — NON-NEGOTIABLE
Preserve the subject’s exact facial structure, facial proportions, skin tone, eye shape, nose shape, lips, hairstyle, hair color, age appearance, skin texture, body proportions, and all recognizable identity traits from the uploaded character reference image.
The subject must clearly remain the same individual.
Do not generate a different face, hairstyle, ethnicity, age, or body type.
Do not beautify, redesign, stylize, or reinterpret the subject’s identity.
JERSEY / WARDROBE LOCK
Accurately recreate the uploaded jersey reference as the subject’s match-day outfit.
Preserve the jersey colors, collar shape, sleeve design, stitching, fabric texture, logos, sponsor marks, club branding, typography, stripe placement, patches, and all visible details.
The jersey must look like a real official football jersey worn naturally in a stadium crowd.
Do not invent a different jersey, change team colors, distort logos, remove sponsor details, or alter branding placement.
SCENE & CONTEXT
The subject is seated casually in stadium chairs, surrounded by real fans, slightly leaned back with a natural candid expression.
The subject looks slightly off-camera as if watching the game or reacting to a moment on the pitch.
The image should feel like a real televised crowd cutaway from a UEFA Champions League or Premier League broadcast.
Avoid cinematic portrait styling, studio lighting, selfie angles, fashion editorial composition, or staged photoshoot energy.
COMPOSITION & CAMERA
Use a long-range professional sports broadcast camera perspective from the audience stands.
Telephoto lens compression.
Imperfect live-broadcast framing.
Slightly off-center subject placement.
Natural TV screenshot-style cropping.
The subject should be recognizable but still integrated into the surrounding crowd.
Avoid perfect portrait framing and shallow beauty-style depth of field.
Keep the scene grounded in authentic stadium broadcast realism.
CROWD & STADIUM DETAILS
Surround the subject with a realistic mixed football audience.
Include fans with different ages, appearances, clothing styles, and reactions.
Some fans wear team jerseys, while others wear jackets, hoodies, scarves, or neutral casual clothing.
Crowd spacing should feel natural and imperfect.
Include realistic stadium seats, steps, railings, barriers, and background spectators.
The crowd should feel alive, slightly chaotic, and unscripted.
BROADCAST REALISM
Add a realistic live TV scoreboard overlay in the top corner.
Include subtle broadcast graphics, live TV compression artifacts, slight digital noise, mild motion softness, natural broadcast sharpness, and slight focus imperfections.
Use realistic stadium lighting and compressed background depth from a telephoto broadcast lens.
The frame should feel like a paused screenshot from live television, not a designed poster.
LIGHTING & COLOR
Use natural stadium lighting with balanced but slightly imperfect broadcast exposure.
Keep realistic skin tones and authentic sports broadcast color grading.
No cinematic color grading.
No glossy commercial lighting.
No artificial studio highlights.
OUTPUT STYLE
Ultra-realistic live football broadcast screenshot.
Documentary-style sports realism.
Authentic televised match atmosphere.
Genuine crowd energy.
Realistic human anatomy.
Real fabric behavior.
Accurate jersey details.
Natural facial expression.
4K live broadcast realism.
Telephoto sports photography look.
The final image should feel like a real screenshot captured during a live football match, not an AI-generated portrait.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 aspect ratio.
Professional social media composition.
Full frame optimized for Instagram portrait format.
Do not generate square, landscape, widescreen, cinematic, 16:9, or cropped compositions.
Text from the author's doc; differs slightly from the site.
Live football broadcast spectator shot
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Accidental snapshots
Full prompt, verbatim · 3,358 characters
Show the full prompt
Create a photorealistic live football television broadcast spectator shot using the uploaded image as the only facial identity reference.
FORMAT LOCK
Vertical 3:4 composition, Authentic sports-broadcast camera framing, Professional live television coverage, Realistic telephoto-lens compression.
IDENTITY LOCK
Preserve the exact face, skin tone, facial proportions, age appearance, hairstyle, eyebrows, eyes, nose, lips, jawline, and recognizable identity from the uploaded image, Do not alter identity, Do not beautify facial features, Do not change ethnicity, gender presentation, or cultural appearance.
SCENE
The subject is attending a FIFA World Cup 2026 match, The subject supports {{Supported Country}}, The subject is located inside a premium VIP hospitality suite overlooking the football pitch, Body naturally turned approximately 45 degrees toward the field, Head follows the same direction, No eye contact with the camera, The subject appears unaware of being filmed.
OUTFIT
Official {{Supported Country}} {{Jersey Style}} national-team jersey, Elegant dark blazer or cardigan, Black trousers, Slim luxury watch, Premium football-event lanyard, Small supporter scarf matching {{Supported Country}} colors.
EXPRESSION
{{Expression}} expression, Watching an important attacking moment, Focused eyes, Slightly tightened eyebrows, Slightly parted lips, Natural emotional tension.
POSE
Both hands resting naturally on the glass railing, Relaxed fingers, No waving, No thumbs-up, No peace signs, No camera-facing gestures.
SURROUNDING CROWD
Five additional spectators positioned naturally around the subject, Mixed supporters of {{Supported Country}} and {{Opponent Country}}, Natural match-day clothing, Everyone focused on the pitch, Partial crops at frame edges, Softly blurred foreground spectators for realistic stadium depth.
ENVIRONMENT
Luxury VIP suite, Subtle glass reflections, Premium stadium architecture, Softly blurred crowd background, Authentic World Cup atmosphere, Natural broadcast-camera perspective.
CAMERA
200mm–400mm telephoto-lens look, Waist-up framing, Sharp face and eyes, Slightly softer surrounding spectators, Natural asymmetrical sports-broadcast composition.
LIGHTING
Realistic stadium lighting, VIP-suite ambient light, Reflected pitch illumination, Natural skin texture, Realistic clothing folds, Mild live-broadcast compression.
SCOREBOARD
Compact professional football-broadcast scoreboard positioned in the top-left corner, Timer: {{Match Minute}}, Score: {{Supported Country}} {{Score}} {{Opponent Country}}
STYLE
Small white rounded timer box, Dark charcoal team boxes, Accurate country flags, Bold three-letter country codes, Bright cyan score boxes, Compact connected broadcast graphics, Clean professional sports-network aesthetic, No LIVE label, Do not cover the subject.
OUTPUT STYLE
Photorealistic sports broadcast, Authentic spectator cutaway shot, Realistic football television coverage, Natural stadium atmosphere, Professional live-camera realism, Ultra-detailed 4K broadcast quality.
NEGATIVE PROMPT
Studio portrait, Fashion photography, CGI appearance, Illustration, Cartoon style, AI beauty filters, Unrealistic hands, Exaggerated facial expressions, Looking at camera, Waving, Low-quality broadcast graphics, Text overlays, Watermarks, Distorted anatomy, Duplicated spectators, Unrealistic lighting.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 4,695 characters
Show the full prompt
If the uploaded photograph depicts me, use it as my personal visual reference. Create a new, highly photorealistic portrait imagining how I might have appeared around 1985. Use the reference to understand my general appearance, facial characteristics, hairstyle, proportions, expression, and overall visual presence, then naturally reinterpret them within an authentic mid-1980s setting. Create a newly composed photograph, not an exact reproduction or reconstruction of the uploaded image.
FORMAT LOCK
Vertical 3:4 portrait, professional photographic composition, natural head-and-shoulders or upper-body framing, natural subject placement, similar portrait orientation to the reference, realistic camera perspective, balanced negative space, natural photographic depth, no unnecessary additional people, no duplicated subjects.
REFERENCE LOCK
Use the uploaded image as the personal appearance reference. Maintain the subject’s general facial proportions, eye characteristics, hair characteristics, skin-tone appearance, facial-hair characteristics where applicable, general expression, body proportions, and overall visual presence. Do not perform forensic facial matching or reproduce the reference photograph exactly. Introduce natural photographic variation so the result feels like an independent photograph of the same subject in another era.
1985 FASHION
Believable mid-1980s everyday fashion: voluminous period-appropriate hairstyle, acid-wash or classic denim, high-waisted jeans, vibrant 1980s top, structured denim or bomber jacket, classic 1980s accessories, period-appropriate eyewear where suitable, and authentic 1980s grooming. Keep the styling realistic rather than exaggerated or costume-like.
HAIR & BEAUTY
Authentic 1985 grooming and beauty aesthetics: natural hair volume, soft textured styling, period-appropriate makeup where suitable, rosy blush, matte lipstick where appropriate, natural-looking skin, and believable 1980s grooming. Do not radically alter the subject’s underlying appearance.
1985 SETTING
Believable mid-1980s interior with subtle background elements such as a cassette player, cassette tapes, vinyl records, analog stereo equipment, vintage posters, wood furniture, period magazines, and warm neutral décor. Keep background elements secondary and slightly out of focus. No smartphones, laptops, modern LED fixtures, contemporary electronics, modern furniture, or futuristic objects.
LIGHTING
Professional mid-1980s portrait lighting: warm studio illumination, soft directional key light, natural facial shadows, gentle highlights, moderate contrast, realistic light falloff, subtle ambient room illumination, and natural depth of field. Avoid modern cinematic lighting or excessive digital effects.
35MM FILM LOOK
Authentic professional 35mm color photograph captured around 1985: fine natural film grain, warm analog color rendition, gentle contrast, soft highlight rolloff, slight optical softness, natural lens characteristics, subtle exposure variation, realistic film texture, and authentic period color response. No heavy scratches, fake dust, extreme fading, or exaggerated old-photo filters.
PHOTOREALISM LOCK
The final image must look like a real professional photograph. Prioritize realistic human anatomy, natural skin texture, individual hair strands, realistic fabric, accurate clothing construction, natural shadows, realistic reflections, authentic lens depth, and believable environmental perspective. Avoid plastic skin, CGI characteristics, artificial facial reconstruction, and excessive AI smoothing.
ORIGINALITY & SIMILARITY SAFETY
Create a fresh photographic interpretation based on the uploaded personal reference. Do not reproduce the original image exactly, copy its pixel arrangement, perform forensic identity matching, reconstruct the original photograph, copy a recognizable third-party person’s likeness, copy another photographer’s distinctive photograph, recreate existing artwork, or mimic a specific living artist’s signature style. Create natural variation in composition, lighting, background arrangement, wardrobe details, and photographic characteristics.
NEGATIVE CONSTRAINTS
No exact image recreation, pixel-for-pixel copying, forensic identity matching, third-party likeness replication, generic replacement face, facial distortion, unnatural beautification, plastic skin, CGI, illustration, painting, cartoon, anime, caricature, cosplay, modern clothing, modern technology, contemporary décor, futuristic objects, excessive film grain, fake scratches, artificial dust, heavy sepia, extreme color grading, excessive blur, HDR appearance, smartphone-camera aesthetics, or exaggerated costume-like 1980s styling.
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Cinematic portraits
Full prompt, verbatim · 2,249 characters
Show the full prompt
Create a highly photorealistic Face Through the Decades portrait timeline showing the same person across 16 different decades, from the 1880s to the 2050s. Use the uploaded image as the only identity reference and preserve the person's recognizable facial identity throughout every panel.
FORMAT
Vertical 3:4 composition. Exactly 16 equal portrait panels in a perfect 4×4 grid.
Row 1: 1880s | 1890s | 1900s | 1910s
Row 2: 1920s | 1930s | 1940s | 1950s
Row 3: 1960s | 1970s | 1980s | 1990s
Row 4: 2000s | 2010s | 2020s | 2050s
Keep all panels perfectly aligned with identical sizes, consistent spacing, matching portrait scale, and readable decade labels centered at the bottom.
IDENTITY & EVOLUTION
Preserve the same facial structure, eyes, nose, lips, jawline, skin tone, hairstyle characteristics, proportions, and overall identity in every decade. Create a natural progression from younger to older, with realistic aging and subtle changes over time.
DECADE STYLE
Automatically adapt clothing, hairstyle, grooming, accessories, background, and photography style to each historical era.
1880s–1910s: Victorian and Edwardian, sepia or antique photography.
1920s–1940s: authentic period fashion, classic black-and-white portraits.
1950s–1960s: post-war and early modern fashion, early color photography.
1970s–1980s: distinctive vintage fashion, warm analog film.
1990s–2000s: minimalist contemporary fashion, late film to early digital photography.
2010s–2020s: modern fashion and high-resolution digital photography.
2050s: believable future version of the same person, naturally older, refined realistic future fashion, grounded and realistic, not sci-fi.
FINAL LOOK
Ultra-realistic historical portrait photography, realistic skin and hair, authentic fashion evolution, natural aging, historically accurate visual styles, each panel with its own photographic character.
The final image must clearly show:
ONE PERSON → 16 DECADES → 1880s TO 2050s
AVOID
Different identities, distorted faces, inconsistent facial features, exaggerated aging, missing or duplicate panels, incorrect decade order, repeated outfits, CGI, illustration, cartoon, anime, plastic skin, unreadable labels, watermarks, logos, cyberpunk, or sci-fi elements.
Aspect Ratio: 3:4
UPLOAD A REFERENCE IMAGE BEFORE GENERATING.
MAIN PROMPT
Create a dramatic vertical Art Deco travel poster featuring [CITY NAME] in the style of a 1930s luxury ocean-liner travel advertisement. Make the landmark immediately recognizable, monumental and collectible, with elegant geometric Deco styling.
FORMAT LOCK
Vertical 3:4 composition. Bold theatrical poster design. Strong diagonal movement from lower left to upper right. Upper third reserved for open sky, the pale cream disc and city name.
STYLE TRANSFER
Landmark-led composition: 70% recognizable architecture, 30% Art Deco stylization. Show the primary landmark from a dramatic low viewpoint, towering upward with converging perspective. Use streamlined monumental forms, geometric arcs, straight verticals, stepped details, speed lines and concentric rays.
DETAIL & COLOR
Primary landmarks: [ONE OR TWO LANDMARKS — describe their distinctive shape and appearance from below]. Secondary elements: [THREE TO FIVE SUPPORTING ELEMENTS, including a bridge, river, canal, road or shoreline creating diagonal movement]. Use [SIX TO EIGHT COOL DESATURATED COLOURS TUNED TO THE CITY], always with deep desaturated navy as the single shadow tone. Blues and teals dominate sky, water and shadows; warm colors only as restrained accents on illuminated planes. Flat printed color planes, museum-print desaturation, subtle even paper grain.
COMPOSITION LOCK
Place a pale cream disc behind the main landmark with clean radiating rays. Keep the disc cool and quiet, never sunset-like. At the upper centre, place only “[CITY NAME]” in uppercase geometric sans-serif with wide letter spacing and a single solid color. Preserve strong diagonal energy and monumental scale.
STYLE
1930s luxury-line Art Deco travel poster, sophisticated geometric design, theatrical hard-edged lighting, flat ink printing, collectible museum-poster aesthetic, clean contemporary finish.
NEGATIVE PROMPT
Yellow, mustard, ochre, gold, amber, orange sky, warm sunset, warm overall cast, soft gradients, photorealism, 3D rendering, neon, cartoon style, fake aging, distressed paper, sepia, brown filters, clutter, extra text, incorrect spelling, additional typography, logos, watermark.
TEXT LOCK
Exact spelling: “[CITY NAME]”. No text other than the city name.
ASPECT RATIO
3:4
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Travel posters
Visible excerpt, verbatim · Copy gives these lines only
Create a premium editorial travel illustration of {{Destination}}, {{Country Region}} using a minimal layered cutout illustration style. The result should feel like a premium Scandinavian editorial po…
FORMAT LOCK
Vertical 3:4 poster composition. Strong foreground, midground, and background separation. Large open breathing space. Refined negative space. Balanced editorial composition. Minimal object density. Do…
STYLE
Same text as the visible prompt except placeholder notation: site uses {{Country Region}}, {{Destination}}, source uses [COUNTRY / REGION], [DESTINATION], [DISTANT NATURAL FEATURE / MOUNTAIN / SKYLINE], [ICONIC LANDMARK / ARCHITECTURAL ANCHOR], [NATURAL FEATURE / WATERFRONT / LANDSCAPE], [PRIMARY COLOR], [SECONDARY COLOR]
Show the variant
Create a premium editorial travel illustration of [DESTINATION], [COUNTRY / REGION] using a minimal layered cutout illustration style. The result should feel like a premium Scandinavian editorial poster, modern travel magazine artwork, sophisticated museum-shop print, and elegant minimalist paper collage, not a busy handmade craft artwork.
FORMAT LOCK
Vertical 3:4 poster composition. Strong foreground, midground, and background separation. Large open breathing space. Refined negative space. Balanced editorial composition. Minimal object density. Do not overcrowd the scene.
STYLE
Minimal layered cutout illustration. Large clean paper shapes. Bold geometric masses. Simplified layered depth. Soft subtle shadows. Clean slightly rounded cut edges. Ultra-simplified shapes. Premium editorial design. No photorealism. No DIY craft aesthetic. No excessive texture.
TEXTURE
Extremely subtle paper grain. Smooth premium paper surface. Almost invisible texture. No heavy fibers, embossed texture, rough paper effects, or handmade imperfections.
FOREGROUND
Create minimal landscape forms representing [NATURAL FEATURE / WATERFRONT / LANDSCAPE]. Use clean abstract terrain and only a few environmental details. Avoid dense vegetation, excessive flowers, or clutter.
MIDGROUND
Feature [ICONIC LANDMARK / ARCHITECTURAL ANCHOR] as the main visual focal point. Surround it with simplified geometric buildings or structures using recognizable silhouettes and minimal window details.
BACKGROUND
Place [DISTANT NATURAL FEATURE / MOUNTAIN / SKYLINE] in the far distance. Add a large clean sun, simple rounded clouds, and spacious atmospheric layers.
ARCHITECTURE
Use simplified isometric-inspired geometric forms. Clean building silhouettes. Minimal windows. Strong recognizable shapes. No unnecessary architectural micro-detail.
COLOR & LIGHTING
Use a restrained 4–6 color palette featuring [PRIMARY COLOR], [SECONDARY COLOR], crisp white, soft slate blue, and pale grey. Vibrant but sophisticated. Bright neutral daylight. Slightly cool-neutral lighting. Fresh whites. No yellow tint or sepia. Soft directional shadows with subtle dimensional depth.
TYPOGRAPHY
Add ”[DESTINATION]” using minimal elegant sans-serif typography. Clean spacing, precise alignment, and understated premium placement. Keep the text perfectly readable.
MOOD & QUALITY
Calm, modern, refined, premium, slightly whimsical, mature high-end travel branding aesthetic. Ultra-clean edges. High resolution. Print-ready quality. Sharp graphic forms. No blur or noise. Sophisticated design-forward travel poster.
NEGATIVE PROMPT
Avoid photorealism, 3D CGI, busy handmade paper craft, DIY collage, excessive paper texture, heavy fibers, rough edges, dense vegetation, excessive landmarks, clutter, warm sepia tones, yellow lighting, gradients, blurry edges, noise, unreadable text, spelling errors, logos, watermarks, or low-quality rendering.
Pastel paper-collage city travel poster
Figure potential: low (flat 2D look)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Travel posters
Full prompt, verbatim · 3,310 characters
Show the full prompt
Create a refined vertical travel-poster illustration of {{City Name}}.
CORE CONCEPT
Depict the city as a calm, elegant pastel paper-collage artwork combining {{Iconic Vehicle}} and {{Primary Landmark}} within one harmonious composition.
The artwork should feel like a premium museum-quality travel poster, collectible souvenir print, handcrafted paper diorama, and sophisticated editorial illustration.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 3:4 poster composition.
Gallery-quality print design.
Premium travel poster layout.
Do not generate landscape, square, widescreen, panoramic, or cropped compositions.
PORTRAIT COMPOSITION
Large open sky occupying approximately 35–45% of the composition.
Balanced negative space.
Clear foreground, midground, and background separation.
Elegant poster hierarchy.
Relaxed, airy composition.
Never overcrowded.
STYLE
Minimal pastel paper-collage illustration.
Layered paper-cut artwork.
Flat geometric color shapes.
Clean silhouettes.
Subtle 3D paper relief.
Visible paper grain.
Soft tactile paper texture.
Gentle paper shadows.
Matte gouache poster finish.
No photorealism.
No CGI.
No glossy materials.
No excessive detail.
FOREGROUND
Place {{Iconic Vehicle}} as the primary foreground subject.
The vehicle should occupy approximately 25–35% of the composition.
Surround it with a small selection of:
Tiny pedestrians.
Cyclists.
Flower planters.
Benches.
Street lamps.
Local urban details.
MIDGROUND
Create a simplified city environment using:
{{Local Architecture}}
{{Local Streetscape}}
{{Local Parks or Waterfront}}
{{Local Cultural Details}}
Keep forms graphic, simplified, and elegant.
BACKGROUND
Feature layered paper-cut silhouettes of:
{{Primary Landmark}}
{{Secondary Landmarks}}
The skyline should fade gradually into soft atmospheric layers.
Landmarks should remain instantly recognizable while staying highly simplified.
SKY
Large calm pastel sky.
2–5 paper clouds.
Small paper birds.
Generous breathing room.
Never busy.
COLOR SYSTEM
Soft sky blue.
Warm cream.
Light beige.
Pale stone gray.
Muted sage green.
Powder mint.
Dusty coral.
Soft peach.
Off-white.
Accent color:
{{Accent Color}}
LIGHTING
Bright soft morning daylight.
Gentle ambient illumination.
Soft shadows.
Clean atmosphere.
Warm optimistic mood.
TYPOGRAPHY
Upper-left corner:
{{City Name}}
Below:
{{Local Language Name}} · {{English Name}}
Subtitle:
{{City Tagline}}
Use elegant white geometric sans-serif typography.
Wide letter spacing.
Minimal hierarchy.
Premium travel-poster design.
OUTPUT STYLE
Premium travel illustration.
Museum-quality city poster.
Collectible souvenir artwork.
Paper-cut diorama aesthetic.
Refined editorial design.
Sophisticated travel advertising.
Pastel paper-collage artwork.
Modern travel-poster design.
FINAL RESULT
A refined pastel paper-collage city poster featuring {{Iconic Vehicle}} and {{Primary Landmark}} as the primary visual anchors, using tactile paper textures, elegant geometric layering, generous negative space, and a sophisticated modern travel-illustration aesthetic.
NEGATIVE PROMPT
Avoid photorealism.
Avoid over-detailed architecture.
Avoid heavy outlines.
Avoid crowded scenes.
Avoid readable advertisements.
Avoid brand logos.
Avoid watermarks.
Avoid neon colors.
Avoid cyberpunk aesthetics.
Avoid dark lighting.
Avoid messy compositions.
Same text as the visible prompt except placeholder notation: site uses {{City}}, source uses [CITY], [LANDMARKS], [LOCAL ELEMENTS]
Show the variant
[CITY] — PREMIUM COLLAGE TRAVEL POSTER
-------------------------------------------------------------------------------------------------
MAIN PROMPT
Create a premium surreal collage-style travel poster of [CITY], combining vintage travel advertising, mid-century editorial collage art, Bauhaus-inspired design, Swiss poster layouts, and contemporary luxury print aesthetics.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 3:4 poster composition.
Gallery-quality print design.
Professional travel poster layout.
Do not generate square, landscape, widescreen, or cropped compositions.
CITY DNA
Visually represent the identity of [CITY] through a curated collection of its most iconic landmarks, architecture, monuments, cultural references, transportation elements, skyline features, and natural surroundings.
The city should be immediately recognizable through its visual storytelling.
MAIN COMPOSITION
Feature these 3–4 iconic landmarks from [CITY]: [LANDMARKS].
Arrange them in a premium layered collage composition.
Mix famous architecture, monuments, skyline elements, cultural symbols, transportation icons, and local nature references.
Allow elements to overlap naturally.
Use mixed scales throughout the composition.
Some landmarks should appear oversized while others remain smaller supporting elements.
Create a floating editorial hierarchy with strong visual balance.
Allow selected landmarks to extend beyond composition boundaries for a more dynamic poster design.
COLLAGE STYLE
Vintage-modern collage aesthetic.
Editorial magazine composition.
Mixed-media paper collage.
Cut-paper photography style.
Minimalist surrealism.
Luxury museum poster aesthetic.
Handcrafted editorial assembly.
High-fashion print campaign energy.
The collage should feel intentionally designed rather than randomly assembled.
LANDMARK TREATMENT
Combine black-and-white cutout photography with selective color accents.
Blend realistic landmark photography with graphic design elements.
Use sharp cutout edges.
Introduce varying scales between landmarks.
Partially crop selected buildings for visual tension.
Merge architecture with abstract geometric graphics.
Allow photographic and graphic elements to coexist naturally.
GRAPHIC DESIGN ELEMENTS
Include abstract geometric circles.
Rectangles.
Minimal lines.
Editorial grids.
Cross marks.
Paper tears.
Halftone textures.
Ink scribbles.
Dotted patterns.
Stamp marks.
Hand-drawn symbols.
Travel-document details.
Editorial layout accents.
Abstract shadow shapes.
Cut-paper layering effects.
These elements should support the composition without overpowering the city landmarks.
COLOR SYSTEM
Warm beige or off-white paper background.
Limited but intentional color palette.
Muted vintage tones combined with selective saturated accents.
Screenprint-inspired color treatment.
Bold geometric color blocks.
Retro-inspired reds, blues, yellows, greens, and oranges used selectively.
Maintain premium visual restraint.
TEXTURE & MATERIALS
Matte paper texture.
Subtle film grain.
Vintage print imperfections.
Museum-quality print surface.
Soft collage shadows.
Paper layering depth.
Fine cutout edges.
Premium tactile print feel.
Avoid glossy digital rendering.
COMPOSITION STYLE
Clean editorial balance.
Large negative space.
Centered visual hierarchy.
Layered depth.
Asymmetrical composition.
Surreal spatial arrangement.
Luxury gallery-poster feeling.
Modern art exhibition aesthetic.
LIGHTING
Soft studio lighting.
Natural paper shadows.
Subtle depth between layers.
Premium print realism.
Balanced contrast.
High-detail rendering.
OPTIONAL LOCAL DETAILS
Integrate these subtle local details from [CITY]: [LOCAL ELEMENTS].
These details should feel like discovered travel artifacts within the poster.
Examples:
Local plants.
Street signs.
Vintage tickets.
Transit maps.
Postmarks.
Travel stamps.
Handwritten notes.
Cultural patterns.
Historic references.
Minimal typography fragments.
OUTPUT STYLE
Luxury travel poster.
Premium editorial collage.
Museum-quality print design.
Vintage-modern visual identity.
High-detail paper collage artwork.
Bauhaus-inspired composition.
Swiss design influence.
Sophisticated travel advertising aesthetic.
Gallery-ready print quality.
Ultra-detailed 3:4 vertical poster.
Colossal cloud animal wildlife landscape
Figure potential: low (photographic scene, not a character form)Prose
By @visualaiclub · licence on the site: attribution · source ↗ · collection: Surreal worlds & scans
Full prompt, verbatim · 3,971 characters
Show the full prompt
PROMPT SETTINGS
Animal: Deer
Use the selected animal for both the colossal cloud-shaped animal and the smaller animals in the foreground.
MAIN PROMPT
Create a surreal cinematic wildlife landscape featuring an enormous colossal cloud-shaped version of the selected animal towering over a vast open grassland.
FORMAT LOCK — NON-NEGOTIABLE
Vertical 4:5 composition.
Epic scale comparison.
The colossal animal must dominate the entire sky and background.
The smaller animals must appear in the foreground.
Do not generate landscape format.
Do not change the environment style.
ANIMAL LOCK — NON-NEGOTIABLE
Use only the selected animal from the prompt settings.
The giant cloud-shaped animal must be the selected animal.
All smaller animals in the foreground must also be the same selected animal.
Do not mix different animal species.
Do not add unrelated animals.
Do not create hybrids or fantasy mutations.
COLOSSAL CLOUD ANIMAL
Create a massive sky-sized animal formed from soft white clouds.
The animal should look like a realistic version of the selected species, but sculpted entirely from dense fluffy cloud formations.
Preserve the animal’s recognizable anatomy, head shape, ears, horns, tusks, mane, trunk, snout, legs, body silhouette, and species-specific features.
The cloud texture should be soft, volumetric, detailed, and realistic.
The creature should feel thousands of feet tall, like a living cloud monument walking across the horizon.
FOREGROUND ANIMALS
Place multiple normal-sized versions of the same selected animal in the grassy foreground.
They should be grazing, walking, resting, or naturally interacting.
The smaller animals should look realistic and proportional.
They must match the selected species exactly.
Their presence should clearly show the enormous scale of the cloud animal above them.
ENVIRONMENT LOCK
Keep the environment fixed as a vast open green grassland with rolling hills.
Wide natural field.
Soft grass texture.
Open plains.
Distant hills.
Subtle dirt paths or natural landscape curves.
No city.
No buildings.
No forest.
No mountains dominating the scene.
No ocean.
No desert.
SKY LOCK
Use a cinematic golden-hour sky with soft blue tones, warm sunlight, and dreamy atmospheric clouds.
The sky should match the reference style: bright, cinematic, warm, peaceful, and epic.
Add soft sunlight from the side.
Add gentle god rays.
Add atmospheric haze.
Add subtle birds flying around the giant cloud animal for scale.
Do not create a night sky, storm sky, dark fantasy sky, sci-fi sky, or colorful surreal sky.
ATMOSPHERE
Soft dust and mist around the feet of the colossal cloud animal.
Gentle haze near the horizon.
Volumetric sunlight.
Dreamlike realism.
Peaceful but epic mood.
Natural wildlife documentary feeling mixed with surreal fantasy scale.
COMPOSITION
Low-angle wide view.
The giant cloud animal fills the upper half of the image.
The normal-sized animals are placed across the lower foreground.
Strong scale contrast.
Balanced cinematic framing.
Natural depth from foreground to horizon.
The image should feel like a real wildlife photograph of a miraculous cloud formation.
LIGHTING
Golden-hour daylight.
Soft warm highlights.
Natural grassland illumination.
Volumetric sun rays.
Gentle shadows.
Realistic atmospheric depth.
No harsh studio lighting.
No artificial neon lighting.
OUTPUT STYLE
Hyper-realistic surreal wildlife landscape.
Colossal cloud animal fantasy realism.
Cinematic grassland photography.
Ultra-detailed cloud texture.
Realistic animal anatomy.
Epic scale.
Dreamy natural atmosphere.
8K quality.
Photorealistic rendering.
National Geographic meets magical realism.
NEGATIVE PROMPT
Avoid mixed animal species, hybrid animals, monsters, fantasy mutations, extra limbs, distorted anatomy, dark stormy sky, night scene, city, buildings, forest, ocean, desert, sci-fi elements, cartoon style, CGI toy look, low detail, blurry textures, text, watermark, logo, and low resolution.
Visual AI Club, visualaiclub.com, every ChatGPT image prompt and its example images as of 2026-09-26; each style's page is linked on its card.
Model facts come from my own research notes on GPT Image 2 and 2.5 (OpenAI's image prompting guide and model docs, LMArena and Artificial Analysis leaderboards, community tests), September 2026.
