AI Art Meme

A social-media meme style retail image built as a clean two-panel comparison. The same blonde woman should appear in both panels wearing a dark red off-shoulder top, with the left side showing her stressed and overwhelmed at a table covered with scattered papers, while the right side shows her smiling warmly as she opens a delivery box full of household shopping items. The composition should feel symmetrical and easy to read, using a simple white background and rounded rectangular panels to create a polished carousel-post aesthetic. Add large blue Portuguese headline text across the top and short captions below each panel to reinforce the emotional contrast between chaos and relief. The delivery box should contain colorful supermarket and home-care products that clearly signal everyday online shopping. The overall image should read like a relatable branded meme about how receiving purchases can instantly improve the mood of a difficult day.

A minimalist social-media concept image showing a wide dusk sky over distant dark hills, designed around a humorous text illusion. The sky should be filled with soft blue-gray clouds and a warm sunset glow near the horizon, while glowing cloud-like letters appear across the middle as if the atmosphere itself were spelling out a short Portuguese phrase. Large bold white text should sit in the upper corner, introducing the image with the tone of a playful observation, as if the signs in the sky are sending an obvious message. The landscape at the bottom should remain simple and dark so the focus stays almost entirely on the sky and the illusion text. The mood should feel casual, meme-ready, and highly shareable, like a post built around a visual coincidence turned into branded humor. The final result should read as a minimalist promotional joke that uses skywriting-style clouds and sunset emptiness to make a short line feel dramatic, funny, and instantly legible.

A candid meme-style portrait set inside a cozy American dive bar, featuring a smiling red-haired woman standing behind the bar and holding two whiskey bottles up at chest level like she is making a joking pre-shift announcement. The background should be packed with liquor bottles, wood-paneled walls, biker-style patches, hats, and roadside-bar decor, giving the scene a very local, informal, working-class nightlife atmosphere. Her expression should feel playful, chaotic, and overly enthusiastic, as if the joke is that the workday has not even started yet and she is already in full party mode. Include bold white meme text near the lower center that reads “Before work 😃” so the image lands clearly as a humorous social-media post built around relatable burnout, service-industry irony, and bar-culture energy.

A meme-style after-work portrait set in the same warm, wood-paneled American dive bar environment, featuring a tired red-haired woman slumped on a bar stool with her arms hanging loosely against the counter and her eyes half closed in pure end-of-shift exhaustion. Her clothes should look visibly damp and disheveled, reinforcing the feeling that the shift has been long, sweaty, chaotic, and physically draining. The bar background should remain detailed and authentic, with liquor bottles, rustic signage, wood textures, and local bar clutter, but the emotional tone should be completely different from a celebratory pose. Add bold white meme text across the center that reads “After work🥺” so the image lands as the exhausted punchline to a before-and-after workplace joke about burnout, service labor, and post-shift collapse.

A humorous fashion-lifestyle image set inside a luxurious walk-in closet, built around the aftermath of choosing the perfect outfit. A stylish blonde woman should stand confidently in the center wearing a polished black one-shoulder top, a sparkling silver skirt, and elegant heels, looking composed and fashion-ready. Behind her, there should be an enormous chaotic pile of discarded clothes, creating a dramatic contrast between the final sophisticated look and the mess left behind during the styling process. The closet should feel rich and upscale, with dark wood cabinetry, warm integrated lighting, neatly displayed shoes, and boutique-like shelving that emphasizes a high-fashion dressing environment. Add a bold Portuguese caption at the bottom that frames the image as the relatable moment right after picking a look. The final image should read like a stylish but self-aware meme about fashion indecision, wardrobe excess, and the absurd amount of effort hidden behind a polished outfit.

[Subject] A deadpan satirical street scene featuring a woman seated at a tiny folding table on the sidewalk, dressed in a teal bathrobe and flip-flops, wearing white oval sunglasses and large hair rollers. She casually smokes a cigarette while sitting behind a handmade cardboard sign that reads “EMOTIONAL ADVICE $2.00,” presenting herself like a pop-up therapist in the middle of urban chaos. [Environment] City street in daylight with multiple burned or burning cars directly behind her, active flames, smoke rising, parked vehicles, storefronts across the road, and a gritty public-street atmosphere that contrasts sharply with the subject’s calm posture. [Composition/Camera] Vertical street photograph, frontal candid framing, subject placed in the lower center foreground at the folding table, empty green chair to the right, burning vehicles dominating the midground and storefront signage in the background, creating a layered meme-like tableau. [Lighting] Natural overcast daylight with muted color grading, soft ambient shadows, orange flame highlights, and smoky haze softening the urban background. [Style/Rendering] Darkly comedic documentary-style image with viral meme energy, realistic street-photo texture, understated cinematic grading, and absurd character styling that makes the scene feel staged but plausible. [Detail constraints] Keep the bathrobe, hair rollers, white sunglasses, cigarette, cardboard advice sign, folding chairs, and burning cars exactly central to the concept; do not add crowds, police action, dramatic motion blur, or polished studio styling.

[Subject]
A split-frame social meme poster showing the same winter-fashion woman recreated in two versions on a snowy alpine slope. In the upper frame she wears a fitted mustard-yellow cold-weather bodysuit with reflective ski goggles and poses confidently toward the camera. In the lower frame the recreated version appears in a black outfit with the same mirrored goggles, wind-swept hair, and a darker mood, creating a humorous before-and-after AI recreation comparison.

[Environment]
Open glacier valley surrounded by rugged snow-covered mountain peaks, pale winter daylight, frozen ground, distant ridge lines, clean white snowfields, no urban elements, isolated outdoor alpine setting.

[Composition/Camera]
Vertical smartphone meme layout with thick black letterbox bars at the top and bottom, two stacked frames separated by a bold horizontal caption strip in Portuguese, medium full-body framing in both panels, centered subject placement, eye-level perspective, symmetrical mountain backdrop, social-media screenshot aesthetic.

[Lighting]
Soft overcast daylight, diffused snow bounce light, gentle facial fill, low-contrast winter atmosphere, slightly brighter exposure in the top frame and moodier exposure in the bottom frame.

[Style/Rendering]
Polished AI-generated fashion realism, cinematic social-media still, crisp clothing contours, realistic snow texture, reflective visor highlights, clean meme composition, editorial outdoor styling, subtle contrast between original and recreated look.

[Detail constraints]
Keep the Portuguese caption band across the middle, preserve the stacked comparison layout, maintain mirrored ski goggles, keep the yellow-vs-black wardrobe contrast, retain snowy mountain continuity between both panels, avoid extra people except a tiny distant silhouette in the lower frame, preserve smartphone meme readability.

Negative prompt
blurry face, malformed hands, duplicate limbs, extra subjects, broken goggles, incorrect mountain perspective, muddy snow texture, unreadable caption, warped body proportions, low-resolution artifacts, overprocessed skin, text glitches, heavy color noise

Suggested parameters
model: SDXL or Flux realism model; aspect ratio: 9:16; stylize: medium; camera: 50mm equivalent; quality: high; sharpness: medium-high; contrast: moderate; seed locked for layout consistency

Delta prompt strategy
1. Lock the two-panel meme structure before changing styling details.
2. Keep the mountain background nearly identical to strengthen the recreation joke.
3. Use wardrobe color contrast as the main storytelling device.
4. Treat the visor reflection as a fixed anchor for identity consistency.
5. Shift only mood and outfit between the upper and lower frame.
6. Keep body pose similar but not perfectly mirrored to feel naturally regenerated.
7. Preserve the central Portuguese caption strip as a meme readability layer.
8. Use snow brightness to separate the subject silhouette from the background.
9. Slightly darken the lower frame so the recreation feels more dramatic.
10. Export with strong vertical framing so it reads instantly as a reel screenshot.

[Subject]
Two women in a theatrical historical-costume tableau, one elderly figure in a long black dress with a large white lace collar and stern expression, one younger figure in a loose brown cloak with a vivid blue headscarf, both isolated as the focal subjects. A bold meme slogan reading WE CHOOSE LOVE appears across the lower center in heavy white letters with a thick red outline.

[Environment]
A hazy stage-like interior filled with purple and blue atmospheric fog, soft silhouettes of background figures, reflective dark floor, nightclub-meets-period-drama energy, cinematic performance space rather than a literal historical room.

[Composition/Camera]
Vertical poster framing, full-body two-character composition, subjects standing side by side with slight spacing, camera placed at chest-to-waist height for a straightforward meme-poster read, centered typography overlay crossing both figures, clean separation from the background.

[Lighting]
Strong purple rim light, cool bluish haze, soft frontal fill on faces and clothing, glowing edge light around both figures, moderate contrast, luminous smoke bloom, dramatic but readable facial detail.

[Style/Rendering]
Stylized cinematic realism, fan-edit poster aesthetic, meme remix energy, glossy social-media graphic finish, vivid color contrast between black dress, brown robe, blue headscarf, and neon-violet background, slightly surreal stage atmosphere.

[Detail constraints]
Keep the lace collar crisp and ornate, preserve the heavy drape of both garments, maintain realistic boots and floor reflections, make the text large and instantly readable, keep the glow around the figures consistent, avoid clutter, avoid extra props, avoid distorted hands or warped typography.

Negative prompt
low resolution, blurry text, unreadable slogan, extra limbs, duplicated people, malformed hands, warped face, muddy clothing folds, flat lighting, washed-out haze, incorrect anatomy, distorted shoes, broken typography, random props, cropped heads, overexposed glow

Suggested parameters
model: SDXL or equivalent, stylize medium, cfg 6.5-8, steps 30-40, aspect ratio 4:5 or 3:4, moderate sharpening, cinematic contrast, typography inpaint pass if needed

Delta prompt strategy
1. Replace the slogan with a different emotional statement while keeping the same composition.
2. Shift the haze from purple-blue to crimson-gold for a warmer dramatic tone.
3. Turn the black dress into a Victorian mourning gown for stronger gothic energy.
4. Swap the brown cloak for a pale ivory robe to increase tonal contrast.
5. Remove the text entirely and use the scene as a pure fan-edit portrait.
6. Push the scene toward vintage theater lighting with amber spotlight and deeper shadows.
7. Convert the image into a cleaner satirical campaign poster with sharper type and flatter background.
8. Add subtle audience silhouettes to strengthen the staged-performance feeling.
9. Reframe closer to a three-quarter crop for stronger facial emphasis.
10. Introduce rain or glitter particles in the haze for a more surreal editorial finish.

GLOBAL LOCK: create a vertical 9:16 art-humor dance reel where a young woman inspired by Vermeer’s Girl with a Pearl Earring is reimagined as a stylish full-body dancer in an opulent museum-like palace hall. Keep her recognizable soft old-master face, pearl earring, and headscarf-inspired styling cues, but adapt the body into a contemporary choreographed performance figure wearing a fitted peach-beige stage outfit that feels elegant, minimal, and fashion-forward. The environment must remain a grand golden interior with glowing chandeliers, ornate walls, polished floors, and warm amber lighting. The camera stays static and centered, framing her full body throughout. No dialogue, no narration, and no lip-sync. The motion is entirely dance-led, with expressive arm lines, small steps, torso turns, and rhythmic pose changes that feel like museum art colliding with modern choreography.

[00:00-00:04.0] Begin with the pearl-earring girl already standing center frame in the ornate hall. She starts moving with small side steps and fluid arm gestures, introducing the surreal concept of a classic painting character joining a dance crew. Keep the facial expression poised and slightly reserved while the body performs modern movement.

[00:04.0-00:08.0] The choreography expands into more deliberate hand placements and angular arm sweeps, with the performer shifting weight from side to side. The warm chandelier light should reflect off the polished floor and highlight the gold decor behind her.

[00:08.0-00:12.5] She turns her shoulders and hips more dynamically, adding a few traveling steps and pose resets while staying centered in the frame. The costume remains fitted and clean-lined, allowing the dance silhouette to read clearly against the ornate hall.

[00:12.5-00:17.0] The movements become bolder and more playful: raised arms, stronger diagonals through the body, and momentary pauses that resemble iconic ad-style dance poses. Even with the choreography, the character should still feel like a living old-master portrait.

[00:17.0-00:21.0] Continue with a sequence of energetic but controlled dance phrases, including sweeping arms, small knee bends, and confident upper-body accents. The hall remains unchanged, with all emphasis on the dancer's silhouette and timing.

[00:21.0-00:24.77] End on a centered finishing phrase with broad open-arm gestures and a balanced final stance. Preserve the pearl-earring identity, the luxurious golden room, and the playful collision of classical painting and contemporary choreography through the last beat.

[Subject] Create a vertical Valentine-style meme card graphic with one stylized female subject composited into the lower-left corner. She appears like a glamorous military pin-up character: young adult woman, long wavy blonde hair, full lips, soft glam makeup, light skin tone, slightly turned face looking off to the side with a detached expression. She wears a tactical helmet, black gloves, and a black strappy tactical vest or armored crop top with matching bottoms. She holds a large black rifle diagonally across her body, the weapon extending from lower-left toward center. Her figure should occupy roughly the lower-left quarter of the canvas and overlap the decorative card border slightly.

[Environment] The overall image is not a realistic scene but a designed greeting-card layout. Background should be pale pink and off-white vertical stripes from top to bottom. In the center place a large ornate scalloped label shape in cream or very light beige with a thick red decorative outline. Add delicate red flourishes, dotted inner border lines, and small doodled red hearts around the text area. The composition should feel like a cute vintage Valentine card mixed with edgy internet meme energy. There is no room interior, no landscape, no photographic background depth beyond the flat poster-like design.

[Composition/Camera] Full-frame vertical poster composition, approximately 4:5. The central ornate label dominates most of the image and leaves a narrow striped margin around all sides. The female subject is cut out and pasted over the lower-left corner, extending slightly upward into the text region but not blocking the main headline. The card includes multiple centered lines of text in the middle. Maintain generous negative space in the center for readable typography. Keep the collage look clean and intentional, like an Instagram carousel meme slide or printable novelty Valentine.

[Lighting] The card background itself is flat and evenly lit like a digital design. The female cutout should retain soft beauty lighting with warm skin tones and enough contrast to separate her from the pale card. No harsh environmental shadows on the card. The subject can have a studio-photo look while the rest of the image stays graphic and flat.

[Style/Rendering] Hybrid style: meme graphic design plus cutout glamour portrait. Use a crisp digital poster finish, smooth pastel palette, and strong contrast between sweet Valentine aesthetics and militarized styling. Dominant colors are blush pink, cream, red, black, and warm blonde. Typography should look playful, rounded, and decorative, like candy-store Valentine lettering. Include the exact center text:
"This Valentine's
card has
been
approved by
Mossad
censorship."

Below that, include smaller handwritten-style lines:
"to:_____"
"from:_____"

Add a couple of tiny doodle hearts near the text. Preserve the intentionally absurd romantic-meets-military meme tone.

[Detail constraints] Do not add extra people. Do not remove the striped background, ornate border, headline text, or to/from lines. Keep one woman only, lower-left placement only, one helmet, one rifle, black tactical outfit, blonde wavy hair, pink-and-cream Valentine card design, centered text, and flat poster-like background consistency.

Negative prompt: extra characters, missing text, misspelled text, incorrect card wording, realistic room background, blood, combat scene, explosions, smoke, duplicated weapon, extra arms, deformed hands, warped helmet, unreadable typography, misplaced subject, dark gothic palette, horror styling, cluttered background, low-resolution meme artifacts, broken border symmetry.

Suggested parameters: aspect ratio 4:5, focal length feel for cutout source 50mm to 85mm portrait, shallow depth of field only on subject source image, high typography fidelity, steps 30-40, CFG/style strength 6-8, sampler DPM++ 2M Karras or equivalent, seed suggestion 214514.

Delta prompt strategy:
1. If the striped background disappears: "restore pale pink and off-white vertical stripes across the full background"
2. If the ornate frame is too plain: "add a large scalloped cream Valentine label with thick red ornamental border and dotted inner decoration"
3. If the woman is too centered: "move the female cutout to the lower-left corner overlapping the border"
4. If the styling becomes too realistic: "keep the image as a meme card / greeting-card poster, not a real-world scene"
5. If the weapon is missing: "place one large black rifle diagonally across the woman's torso"
6. If the text is unreadable: "use clean centered decorative pink-red lettering with exact card wording preserved"
7. If the tone gets too violent: "make the mood ironic and playful, mixing cute Valentine design with edgy internet humor"
8. If the palette shifts dark: "return to blush pink, cream, red accents, black wardrobe, warm blonde hair"
9. If extra props appear: "remove all additional objects except small hearts, border ornaments, and the one rifle"
10. If the bottom labels are absent: "include handwritten-style 'to:_____' and 'from:_____' lines below the headline"

Create a square, meme-like Valentine card on a bright pink gradient background with a soft glowing heart centered behind the subject. Place a seated young woman in the foreground, photographed full body, wearing a military helmet with goggles and tan tactical gear styled like a playful pin-up costume. Keep her pose relaxed and slightly tilted, with one leg folded and the other extended, facing the camera with a mildly teasing expression. Add large bubble-style script text across the top in white with a hot-pink outline, plus handwritten "to:" and "from:" lines in the lower left so the image reads like a shareable greeting card. The overall look should feel glossy, ironic, flirtatious, and optimized for reposting as a social-media graphic.

[Subject] One chibi penguin mascot standing on a wooden outdoor bench, centered lower-middle, with large blue eyes, open orange beak, and yellow plaid scarf. Body is fluffy gray-and-white with small orange feet visible.

[Environment] Elevated city-view park setting in daylight. Tree branches frame upper-right area, urban skyline visible in distance including a tall landmark tower, and greenery in middle ground.

[Text overlay] White text near upper center reading "Have a very".

[Composition/Camera] Vertical social-card format. Mascot occupies central-lower area, text at top, skyline and trees form contextual backdrop. Bench slats provide base line at bottom.

[Lighting] Bright clear daylight with crisp shadows and vibrant sky-blue tones.

[Style/Rendering] 3D cartoon mascot composited over real city-park photo, greeting/meme format for social platforms.

[Detail constraints] Keep single centered mascot on bench, top white text "Have a very", and city-overlook park background with tree framing. Do not add extra characters, additional text lines, or dark cinematic filters.

Negative prompt: wildlife realistic penguin, indoor scene, nighttime city lights, crowded people, long text paragraphs, watermark, glitch effects, angry expression.

Suggested parameters: aspect ratio 9:16 vertical; steps 22-34; CFG/style strength 6-8; sampler DPM++ 2M Karras; seed 31478250.

Delta prompt strategy (top drift risks + corrective micro-prompts):
1) If text disappears -> "add top-center white text 'Have a very'".
2) If scarf missing -> "yellow plaid scarf with pom-pom ends".
3) If mascot not on bench -> "penguin standing on wooden bench slats".
4) If city background changes -> "daytime city skyline seen from leafy park overlook".
5) If composition shifts -> "mascot centered lower-middle with top text headroom".
6) If style turns realistic -> "cute 3D chibi mascot style".
7) If extra elements appear -> "single mascot, minimal clutter".
8) If mood darkens -> "bright cheerful daylight ambiance".
9) If character shrinks -> "large foreground mascot focus".
10) If tree framing missing -> "include leafy branches in upper-right area".

HIGH-GRANULARITY VISUAL INVENTORY
- Subject(s): one stylized baby penguin character (cartoon/3D look), centered in foreground, slightly tilted head, large glossy blue eyes, small orange beak, rounded soft body proportions, black-and-white fur texture, tiny orange feet, friendly curious expression.
- Clothing & materials: yellow knitted scarf wrapped around neck with visible woven pattern and fabric folds; no other garments.
- Props/objects: wooden slat bench in foreground where character stands; curved metallic structure at left edge (partial circular/elliptical reflective surface); partial chair/bench arm detail at bottom-right edge.
- Environment: outdoor elevated viewpoint in a city park. Background includes trees, shrubs, distant skyline, and a tall landmark skyscraper (needle-like top, resembling Taipei 101 silhouette) under bright blue sky with light clouds. Daytime clear weather.
- Composition: vertical portrait (9:16). Character occupies lower-middle central area. Text overlay near top center reads "Have a very" in white letters with dark outline. Background is visible enough for place context but less dominant than character.
- Lighting: natural daylight from above/front-left, soft shadows, bright ambient fill. Character lighting is blended to match outdoor scene with gentle highlights on eyes and scarf.
- Color palette: sky blue and leaf green environment; black/white/orange penguin tones; yellow scarf as key accent. Medium contrast, cheerful saturation.
- Image style: hybrid composite (3D cartoon character inserted into real photo background), social meme/reel cover aesthetic, clean edges, playful tone, high readability for mobile.

MASTER PROMPT
[Subject] A cute stylized baby penguin character with a rounded body, large shiny blue eyes, small orange beak, and tiny orange feet, wearing a textured yellow knitted scarf, standing on a wooden bench, head slightly tilted with a gentle inquisitive expression.

[Environment] Outdoor urban park overlook in daylight, trees framing the scene, distant city skyline and a tall landmark tower in the background, clear blue sky with light clouds, natural greenery in mid-ground.

[Composition/Camera] Vertical 9:16 framing, penguin centered in lower-middle foreground, partial metallic curved structure on the left edge, bench slats visible at bottom, skyline layered in the distance, social-post style framing with room for headline text at top.

[Lighting] Soft natural daytime illumination, even ambient fill, mild highlights on eyes/beak/scarf, subtle realistic contact shadows on bench, balanced exposure to keep both character and skyline readable.

[Style/Rendering] High-quality 3D-cartoon character composited into a real photographic background, playful wholesome mood, clean edges, crisp mobile-first visual clarity, meme-cover readability.

[Detail constraints] Keep exactly one penguin, keep yellow scarf, keep wooden bench foreground and city+tower background, retain bright blue-sky park context, keep top text area with short phrase styling, do not add extra characters or holiday props.

NEGATIVE PROMPT
extra animals, multiple penguins, humans in foreground, winter snow scene, night city, rain, blurry mascot, distorted face, oversized beak, missing scarf, different scarf color, crowded text blocks, watermark, logos, neon cyberpunk color grading, heavy film grain, painterly brush strokes, low-res artifacts.

SUGGESTED PARAMETERS
- Aspect ratio: 9:16
- Lens feel: 28-35mm smartphone-equivalent environmental portrait
- Depth of field: moderate (subject crisp, background slightly soft but recognizable)
- Steps: 28-36
- CFG / guidance: 5.5-7.0
- Sampler: DPM++ SDE Karras or DPM++ 2M Karras
- Style strength: medium (preserve realism in background and stylization in character)
- Seed suggestion: 731905 as baseline, iterate by +/- 30

DELTA PROMPT STRATEGY (TOP 10 DRIFT FIXES)
1) If character looks too realistic:
   - "stylized 3D cartoon baby penguin, soft toy-like proportions, oversized glossy eyes"
2) If scarf color changes:
   - "yellow knitted scarf with visible weave texture wrapped once around neck"
3) If skyline disappears:
   - "distant urban skyline with a tall needle-topped landmark tower in background"
4) If scene becomes indoor:
   - "outdoor park overlook, trees, daylight sky, city visible in the distance"
5) If bench is missing:
   - "penguin standing on wooden slat bench in foreground"
6) If text is wrong position:
   - "short headline text at top center: 'Have a very', white letters with dark outline"
7) If lighting mismatches composite:
   - "natural midday sunlight with soft ambient fill and grounded contact shadow"
8) If extra subjects appear:
   - "single mascot character only, no additional animals or people"
9) If mood becomes dark/sad:
   - "bright cheerful wholesome social-post mood, clean vibrant colors"
10) If edges look rough:
   - "clean compositing edges, high clarity, no jagged cutout artifacts"

[Subject] One chibi-style baby penguin mascot standing upright in center foreground. Round body, black-and-white feather pattern, sleepy/sad half-closed eyes, small orange beak, tiny orange feet, and a yellow scarf around the neck. Expression conveys longing and melancholy.

[Environment] Bright autumn park scene with vivid red-orange trees, green grass, scattered fallen leaves, and a winding path in the background. Golden sky with visible sun near upper center.

[Composition/Camera] Vertical 9:16 meme composition, full-body penguin centered and large in frame. Caption text at top center in bold white with dark stroke: "Mama I miss you". Background depth slightly soft but still readable.

[Lighting] Warm sunset/golden-hour lighting, strong orange cast in sky and trees, soft shadows on ground. Character lit evenly with subtle rim light.

[Style/Rendering] 3D cartoon meme aesthetic with high-contrast cute mascot design, bright saturated autumn palette, social short-video still look.

[Detail constraints] Keep one penguin with yellow scarf, autumn park background, and exact top caption. No extra characters, no logos, no additional text blocks.

Negative prompt: multiple penguins, winter snow scene, indoor room, missing scarf, happy excited expression, blurry caption text, photoreal bird anatomy, watermark overlays, dark nighttime palette.

Suggested parameters: --ar 9:16, focal feel 35mm equivalent, aperture look f/4-f/5.6, steps 26-36, CFG/style strength 4.5-6.5, sampler DPM++ SDE Karras, seed 512906.

Delta prompt strategy (top drift risks + corrective micro-prompts):
1) Caption changes -> "top text exactly: 'Mama I miss you'"
2) Trees lose autumn color -> "vivid red-orange fall foliage under golden sky"
3) Scarf disappears -> "yellow scarf clearly wrapped around neck"
4) Emotion too cheerful -> "sad sleepy eyes and wistful mouth"
5) Character too small -> "penguin occupies large central foreground"
6) Background wrong season -> "autumn park with fallen leaves and warm sun"
7) Extra characters appear -> "single penguin only"
8) Lighting too cool -> "warm golden sunset tone"
9) Body proportions alter -> "round chibi penguin with tiny feet"
10) Text readability poor -> "bold white caption with dark outline"

HIGH-GRANULARITY VISUAL INVENTORY

Subject(s)
- Count: 1 character (baby penguin / penguin chick).
- Proportions: chibi/animated; oversized head; small body.
- Pose: standing upright, centered, facing camera.
- Expression: supportive/concerned; slight frown; eyes wide and glossy.
- Facial details: white fluffy face and belly; dark charcoal/black crown and sides; small orange beak.
- Eyes: large glossy teal-blue eyes with strong highlights.

Clothing & materials
- Scarf: golden-yellow knitted scarf wrapped around neck, ends crossed in front.

Props/objects
- On-image text overlay at top center: “If you need me” in white bold sans-serif with black outline/shadow.

Environment
- Beach/coast at sunset.
- Background: calm ocean with flat horizon; sun low and glowing; warm sky gradient.
- Foreground: sandy/rocky ground near shoreline with small shrubs/bushes on both sides.

Composition
- Aspect: vertical (9:16).
- Framing: full body; penguin centered in lower half; sky and sea behind.
- Camera: eye-level.
- Depth: background slightly softened; subject sharp.
- Negative space: upper sky reserved for text.

Lighting
- Warm sunset backlight from sun over water; gentle rim on character.
- Soft fill from sky; no harsh shadows.

Color palette
- Dominant: warm gold/orange sunset + muted blue-grey sea.
- Accents: yellow scarf, orange beak/feet, teal eyes.

Image style
- Stylized 3D character composited over a photographic sunset beach background; sentimental meme-card style.


MASTER PROMPT (English)

[Subject]
One adorable chibi baby penguin standing centered and facing the camera, oversized head and small body; fluffy white face and belly with charcoal/black head and sides; small orange beak; huge glossy teal-blue eyes with strong highlights; slightly concerned supportive expression.

[Clothing]
Golden-yellow ribbed knitted scarf wrapped around the neck, ends crossed in front.

[Environment]
Sunset beach scene with calm ocean and a low sun near the horizon, warm sky gradient; sandy/rocky foreground near shoreline with small shrubs on both sides.

[Composition/Camera]
Vertical 9:16, full-body centered; penguin in the lower half; sea and sky behind; keep upper sky clean for text; eye-level camera; slight background softness.

[Lighting]
Warm sunset backlight from the sun over the water, subtle rim light, soft sky fill; sentimental mood.

[Style/Rendering]
Stylized 3D animated character composited onto a photographic beach background, clean shading, crisp subject.

[Text overlay]
Add top-centered overlay text exactly: "If you need me" in bold white sans-serif with black outline/drop shadow.

[Detail constraints]
One penguin only; keep the yellow scarf; keep sunset ocean background with visible sun; keep the exact text wording; keep centered composition and soft sentimental lighting.


NEGATIVE PROMPT
extra characters, multiple animals, people, watermark, logo, random text, misspelled phrase, heavy blur, noisy grain, horror, photoreal penguin, deformed beak, asymmetrical eyes, over-saturated neon


SUGGESTED PARAMETERS (for reproducibility)
- Model: SDXL (3D/animation style)
- Aspect ratio: 9:16
- Steps: 30–40
- Sampler: DPM++ 2M Karras
- CFG / guidance: 5.5–7
- Resolution: 1024x1824 (or 1080x1920)
- Seed: 25122025


DELTA PROMPT STRATEGY (top drift points + corrective micro-prompts)
1) Text changes → “top text EXACTLY: If you need me, bold white sans-serif, black outline”
2) Sun/sunset disappears → “low sun near horizon over calm ocean, warm sunset gradient”
3) Scarf changes → “golden-yellow knitted scarf, ends crossed in front”
4) Eyes lose gloss → “huge glossy teal-blue eyes with strong highlights”
5) Expression changes → “supportive concerned expression, slight frown”
6) Background becomes busy → “simple beach foreground, minimal shrubs, clean horizon”
7) Style turns photoreal → “stylized 3D animated penguin, clean shading”
8) Subject not centered → “penguin centered full body, lower half of frame”
9) Lighting becomes harsh → “soft backlit sunset, gentle rim, soft fill”
10) Colors oversaturate → “natural sunset tones, restrained saturation”

[Subject]
Vertical emotional-support 3D scene with a single chibi penguin character standing on a bed indoors at night. Character has fluffy black-and-white body, big blue eyes, orange beak, orange feet, and a mustard-yellow scarf tied around neck.

[Props/objects]
Top-center text reads "I pray" in bold black letters with white outline. Background shows a large window with blue night sky and faint rain/speckled texture. Curtains frame the right side; room is softly lit in cool blue tones. Bed surface visible in foreground.

[Environment]
Cozy bedroom interior at night, calm contemplative atmosphere.

[Composition/Camera]
Vertical 9:16 frame. Penguin centered and full body visible in middle-lower frame. Text placed in upper-middle over window area. Symmetrical window framing supports quiet mood.

[Lighting]
Soft cool moonlight-style illumination from window behind plus gentle front fill on character. Low contrast, serene nighttime ambience.

[Style/Rendering]
3D animated mascot aesthetic with comforting bedtime tone, family-friendly social post design, soft and heartfelt.

[Detail constraints]
Keep one single penguin with yellow scarf, bedroom-night window background, and exact text "I pray". Preserve gentle blue nighttime palette and contemplative mood. Do not add extra characters or bright daytime lighting.

Negative prompt:
daylight room, outdoor landscape, multiple mascots, horror dark shadows, missing text, no scarf, heavy glitch effects, watermark, realistic penguin photography, cluttered bedroom mess

Suggested parameters:
- aspect ratio: 9:16
- lens/focal length: 35mm equivalent
- depth of field: moderate (character sharp, room softly clear)
- steps: 24-36
- CFG/style strength: 5.5-7
- sampler: DPM++ 2M Karras
- seed suggestion: 139680245

Delta prompt strategy (top drift risks and corrective micro-prompts):
1) Drift: text altered
   Fix: "top text exactly: I pray"
2) Drift: scene becomes daytime
   Fix: "nighttime blue-lit bedroom with dark window"
3) Drift: scarf omitted
   Fix: "mustard-yellow scarf tied at neck"
4) Drift: extra characters appear
   Fix: "single penguin character only"
5) Drift: mood becomes energetic
   Fix: "quiet contemplative bedtime atmosphere"
6) Drift: character not centered
   Fix: "penguin centered, full body visible"
7) Drift: window context lost
   Fix: "large window behind with cool blue night sky"
8) Drift: lighting too harsh
   Fix: "soft moonlight and gentle front fill"
9) Drift: proportions become realistic
   Fix: "chibi 3D penguin with oversized head and eyes"
10) Drift: frame horizontal
    Fix: "vertical 9:16 story composition"