Showcase

Image 1 role

Anchor the retriever's face, golden coat, dark eyes, ear shape, muzzle, and body proportions.

Image 2 role

Supply the spacesuit construction, cool cabin light, window shape, and weightless atmosphere.

Prompt goal

Fit the astronaut concept naturally around the animal while keeping the eyes and muzzle clear and the pet as the only hero subject.

Generated result

The result is a newly generated cabin portrait with coordinated costume, light, and pose rather than a cut-and-paste edit.

Review checklist

Face and coat identity; Four plausible paws; Costume edges around fur

Common failure

Suit seams can blend into fur, hide the ears, or change the muzzle when the costume request carries too much detail.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated with OpenAI; not a customer result, live benchmark, or repeatability promise. Provenance: pet-astronaut-triptych.

Image 1 role

Preserve the cat's tuxedo markings, green eyes, whiskers, upright ears, compact scale, and seated posture.

Image 2 role

Guide the glasshouse structure, filtered sunlight, leaf density, stone floor, and humid green palette.

Prompt goal

Place the cat naturally inside the greenhouse with believable ground contact and keep plants behind, not across, the face.

Generated result

The generated scene connects the cat to the floor and surrounding foliage while using the second image only as environmental direction.

Review checklist

Chest marking continuity; Whisker and ear edges; Scale against the floor

Common failure

Dense leaves may cross the face, duplicate whiskers, or make the animal too large for the architectural perspective.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration created with OpenAI; not a user commission, customer result, or benchmark. Provenance: pet-greenhouse-triptych.

Image 1 role

Anchor the corgi's white facial blaze, upright ears, amber coat, short legs, expression, and sturdy body shape.

Image 2 role

Provide the timber cabin, snowy window, woven blanket, warm firelight, and restrained winter palette.

Prompt goal

Create a cozy indoor portrait with the corgi resting near the window, realistic leg length, and no clothing over the face.

Generated result

The generated composition matches warm interior light to the coat and gives the dog convincing contact with the blanket and floor.

Review checklist

Facial blaze; Leg length and paw count; Warm and cool light balance

Common failure

The model may lengthen the legs, mirror the facial blaze, or add decorative winter clothing that hides identifying fur.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated for this site; output details can vary and require visual review. Provenance: pet-corgi-alpine-triptych.

Image 1 role

Preserve the greyhound's narrow face, rose ears, deep chest, long limbs, pale coat, and calm standing pose.

Image 2 role

Contribute wind-shaped grass, pale cliffs, overcast side light, muted fabric tones, and open coastal framing.

Prompt goal

Build an elegant full-body portrait that keeps breed anatomy accurate and uses the coast as context rather than visual clutter.

Generated result

The result is a new editorial photograph with coordinated wind, posture, and light rather than a literal transfer of background pixels.

Review checklist

Long-limbed anatomy; Ear and face shape; Ground contact on grass

Common failure

Long legs may bend unnaturally, paws may disappear in grass, or the face may broaden into a different breed.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated for explanation; not a customer photograph or breed-accuracy benchmark. Provenance: pet-greyhound-coast-triptych.

Image 1 role

Keep the rabbit's upright ears, gray-brown coat, white nose, dark eyes, compact body, and natural crouched posture.

Image 2 role

Guide the low wildflowers, dawn haze, backlight, shallow depth of field, and soft spring color palette.

Prompt goal

Place the rabbit among low flowers without enlarging it, hiding its feet, or letting stems cross its eyes and nose.

Generated result

The generated image uses atmospheric light from the meadow and keeps the animal as a small, readable focal point.

Review checklist

Ear count and shape; Rabbit-to-flower scale; Clear eyes and feet

Common failure

Flowers can become oversized, ears can duplicate, or the animal can take on generic storybook anatomy instead of the source traits.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration; it explains reference roles and does not promise the same animal identity on every run. Provenance: pet-rabbit-meadow-triptych.

Image 1 role

Anchor the parrot's red crown, blue wing feathers, curved beak, eye ring, feet, scale, and alert posture.

Image 2 role

Supply iron greenhouse arches, tropical leaves, humid window light, a natural branch, and layered green depth.

Prompt goal

Create a conservatory portrait with the bird clearly separated from foliage and both feet gripping one plausible perch.

Generated result

The generated composition coordinates feather highlights with window light and builds a new branch-and-glass setting around the subject.

Review checklist

Feather color pattern; Beak and eye ring; Two feet on one perch

Common failure

Bright leaves may merge with wing edges, toes may duplicate, or feather colors may be redistributed across the body.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration; not a wildlife photograph, customer result, or claim of zoological precision. Provenance: pet-parrot-conservatory-triptych.

Image 1 role

Preserve the fictional adult subject's facial proportions, brown eyes, dark hair, skin texture, and calm expression.

Image 2 role

Guide the mountain architecture, mist, travel clothing, cool ambient color, and warm rim light.

Prompt goal

Create a head-and-shoulders fantasy avatar with a readable face and an original world rather than copying a franchise.

Generated result

The output is a new portrait that coordinates clothing and background light while keeping identity cues available for review.

Review checklist

Facial proportions; Hairline and eye color; Original clothing and symbols

Common failure

Heavy world-building can change the face, add franchise-like symbols, or turn a profile crop into an inconsistent wider pose.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated with OpenAI; not a real customer, celebrity, or identity benchmark. Provenance: portrait-fantasy-triptych.

Image 1 role

Anchor the fictional adult face, short dark hair, jaw line, natural skin texture, expression, and head-and-shoulders crop.

Image 2 role

Provide cobalt geometry, blue edge light, dark negative space, and a clean technology-oriented background rhythm.

Prompt goal

Build a square profile avatar whose face remains primary while the second reference contributes only lighting and backdrop structure.

Generated result

The generated avatar changes light and background without treating the geometric reference as another person or scene to copy.

Review checklist

Face and jaw line; Natural skin texture; Clean square crop

Common failure

Strong colored light can flatten skin tone, alter eye color, or let geometric shapes intersect the face and hair.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration; not a celebrity, user commission, customer result, or cross-model benchmark. Provenance: avatar-cobalt-triptych.

Image 1 role

Preserve the adult subject's face, curly hair, freckles, expression, shoulder line, and ordinary wardrobe proportions.

Image 2 role

Guide the warm plaster backdrop, large softbox direction, amber-neutral palette, and clean editorial negative space.

Prompt goal

Create a polished but natural profile portrait with realistic skin and no invented jewelry, text, or brand marks.

Generated result

The result is a newly generated studio portrait with softer background separation and lighting derived from the second image.

Review checklist

Freckles and face shape; Hair edge detail; Profile-ready crop

Common failure

Editorial polish can over-smooth skin, change small facial marks, or add accessories that were absent from the portrait.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration generated for this site; it is not a real client's profile photograph. Provenance: avatar-warm-studio-triptych.

Image 1 role

Anchor the fictional adult's face, silver-streaked hair, brown eyes, glasses, expression, and upper-body proportions.

Image 2 role

Contribute curved towers, sunset haze, warm metal surfaces, teal transit light, and a wide elevated viewpoint.

Prompt goal

Create an original city avatar with compatible perspective and rim light while keeping the face unobstructed and glasses plausible.

Generated result

The generated portrait integrates skyline light into hair and clothing and invents a coherent scene rather than copying a property.

Review checklist

Glasses geometry; Eye and hair identity; Original architecture

Common failure

Reflections can hide the eyes, towers can resemble familiar franchises, or metallic clothing can overwhelm facial identity.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original environment; not a franchise image, celebrity likeness, or customer commission. Provenance: avatar-retrofuture-triptych.

Image 1 role

Preserve the fictional adult's oval face, braided hair, dark eyes, expression, neckline, and simple centered pose.

Image 2 role

Provide layered paper leaves, visible cut edges, coral and teal shapes, shallow shadows, and circular framing.

Prompt goal

Create an original square avatar that translates material and palette without turning the subject into a flat anonymous silhouette.

Generated result

The output combines a recognizable illustrated face with layered paper forms and a new botanical frame around the shoulders.

Review checklist

Face remains distinct; Layered paper edges; No named-style imitation

Common failure

Heavy simplification can remove facial distinctions, merge braids with leaves, or create unreadable paper fragments across the face.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration; the visual direction is generic and does not imitate a named living artist. Provenance: avatar-paper-botanical-triptych.

Image 1 role

Anchor the fictional adult's angular face, close-cropped hair, expression, nose, eye spacing, and dark jacket silhouette.

Image 2 role

Guide one hard side light, window-blind shadows, charcoal tones, restrained grain, and deep negative space.

Prompt goal

Create a dramatic profile portrait with both eyes readable, no smoking props, no text, and no reference to a named film.

Generated result

The new portrait uses controlled shadow and grain while preserving enough midtone detail to evaluate identity and clothing edges.

Review checklist

Eyes remain visible; Face shape in shadow; Clean jacket silhouette

Common failure

Hard shadow can erase one side of the face, change age cues, or introduce props and period details not requested.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration; not a celebrity likeness, film still, customer result, or performance benchmark. Provenance: avatar-noir-light-triptych.

Image 1 role

Preserve the unbranded bottle's amber glass, pump, cylindrical silhouette, material highlights, scale, and blank label layout.

Image 2 role

Guide the pale stone surface, neutral interior, window direction, soft shadow, and restrained lifestyle palette.

Prompt goal

Place one hero bottle naturally on the surface with plausible contact shadow and no invented product wording or claims.

Generated result

The generated campaign image matches bottle reflections to the room and creates a new composition around the source object.

Review checklist

Bottle and pump geometry; Contact shadow; No invented label text

Common failure

The pump can bend, the bottle can float, or the model can invent a label that looks factual but is not verified.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated with OpenAI; labels and product claims are not benchmarked or commercially verified. Provenance: product-lifestyle-triptych.

Image 1 role

Anchor the jar's squat silhouette, matte ceramic body, lid seam, neutral color, blank face, and small physical scale.

Image 2 role

Supply limestone texture, low camera height, hard afternoon sun, distant water, and warm outdoor reflections.

Prompt goal

Create one outdoor hero shot with the jar resting on stone and no invented logo, ingredients, measurement, or claim.

Generated result

The generated image coordinates shadow direction and material color with the terrace while preserving a reviewable product outline.

Review checklist

Jar and lid geometry; Stone contact; Blank packaging surface

Common failure

Hard light can reshape the lid, add a false glossy finish, or make the small jar look like large architectural decor.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration; not a brand asset, customer result, catalog record, or commercial accuracy benchmark. Provenance: product-terrace-triptych.

Image 1 role

Preserve the unbranded shoe's side profile, white midsole, blue mesh, lace count, heel height, and material boundaries.

Image 2 role

Guide the track curve, low dawn light, lane texture, slight motion atmosphere, and ground-level camera perspective.

Prompt goal

Create a grounded hero shot with one shoe, readable construction, realistic shadow, and no athlete or invented brand mark.

Generated result

The result places the shoe on a plausible lane and uses the reference light to reveal mesh and sole texture.

Review checklist

Sole and lace geometry; One complete shoe; Ground contact and scale

Common failure

Laces can multiply, the sole can warp, or a familiar-looking logo can appear even when the source is unbranded.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned unbranded synthetic product demonstration; not a retail item, endorsement, or claim of catalog accuracy. Provenance: product-running-shoe-track-triptych.

Image 1 role

Anchor the kraft bag's gusset, sealed top, rectangular label area, matte paper texture, proportions, and blank branding.

Image 2 role

Provide the stone counter, morning window light, coffee tools, warm wood, and shallow-focus kitchen context.

Prompt goal

Create one upright hero package with secondary props behind it and no invented roast name, certification, weight, or logo.

Generated result

The generated composition integrates paper texture and soft side light while keeping the package as the single focal product.

Review checklist

Top seal and gusset; Blank label area; Readable product silhouette

Common failure

Packages can become asymmetrical, label-like nonsense text can appear, or nearby cups can duplicate and compete with the product.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned unbranded demonstration; all packaging facts remain fictional and must not be treated as a real listing. Provenance: product-coffee-kitchen-triptych.

Image 1 role

Preserve the round blank dial, two simple hands, brushed case, dark strap, lugs, crown, and balanced proportions.

Image 2 role

Guide the dark acrylic surface, narrow blue highlight, controlled reflection, curved backdrop, and premium studio contrast.

Prompt goal

Create one centered watch campaign image with plausible reflections and no logo, numbers, extra hands, or written claim.

Generated result

The output uses the studio reference for light and surface while rebuilding the watch in a newly generated composition.

Review checklist

Dial and hand count; Strap-to-case joints; Reflection matches geometry

Common failure

Small circular details can drift, watch hands can multiply, and reflections can describe a different object than the visible case.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned unbranded synthetic demonstration; not a functional watch design, specification sheet, or retail photograph. Provenance: product-watch-studio-triptych.

Image 1 role

Anchor the clear glass cylinder, cream wax level, centered wick, small flame, blank label area, and wooden lid.

Image 2 role

Supply the rainy window, book-lined nook, soft gray daylight, warm table surface, and quiet evening atmosphere.

Prompt goal

Create a safe-looking tabletop scene with one candle, controlled flame, realistic glass reflection, and no readable branding.

Generated result

The generated image balances cool window light with the flame and places the vessel naturally beside secondary reading props.

Review checklist

Glass and wax geometry; Single wick and flame; No unsafe nearby material

Common failure

The flame can duplicate, glass edges can melt into the background, or books and fabric can appear implausibly close.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned unbranded demonstration; not a real product, fire-safety test, customer image, or regulated marketing asset. Provenance: product-candle-rain-triptych.

Image 1 role

Preserve the fictional adult's face, hair, body type, stance, expression, and practical explorer starting point.

Image 2 role

Provide alpine terrain, luminous plants, cool mist, evening palette, and environmental light from the original valley.

Prompt goal

Create an original ranger with restrained equipment and lighting that belongs in the valley while keeping the face clear.

Generated result

The new full-body composition coordinates clothing, terrain, and atmosphere while retaining identity cues from the first image.

Review checklist

Adult facial identity; Hands and equipment; Original symbols and costume

Common failure

Dense gear prompts can duplicate straps, tools, fingers, or symbols and make the result resemble familiar franchise design.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated with OpenAI; outputs vary and this is not a live benchmark. Provenance: fantasy-ranger-triptych.

Image 1 role

Preserve the fictional adult's face, dark curls, brown eyes, expression, shoulder width, and visible age cues.

Image 2 role

Guide wet stone towers, storm light, wind direction, slate palette, practical cloak material, and guardian atmosphere.

Prompt goal

Create an original citadel guardian with an unobstructed face, restrained gear, and no recognizable franchise emblem.

Generated result

The generated portrait matches environmental highlights across hair and clothing and invents a new guardian composition.

Review checklist

Face and age cues; Cloak and hand anatomy; Original architecture and symbols

Common failure

Armor can hide the subject, lightning can disagree with face shadows, or decorative emblems can resemble existing properties.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration; not a franchise, user commission, customer result, or cross-model benchmark. Provenance: fantasy-citadel-triptych.

Image 1 role

Anchor the adult navigator's face, cropped silver hair, lean build, steady expression, stance, and weathered travel clothing.

Image 2 role

Provide brass instruments, cliff architecture, star-filled sky, strong crosswind, navy palette, and moonlit edge light.

Prompt goal

Create an original navigator using one plausible instrument with clear hands, practical layers, and no named-world iconography.

Generated result

The generated scene links the character's pose to the observatory and uses wind and moonlight across clothing and hair.

Review checklist

Face and hair identity; Hands on one instrument; Original astronomical symbols

Common failure

Complex instruments can create extra hands and unreadable markings, while wind effects can distort clothing or facial features.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original-world demonstration; not a franchise character, customer commission, or benchmark. Provenance: fantasy-observatory-triptych.

Image 1 role

Preserve the fictional adult's round face, short dark hair, freckles, stocky build, focused expression, and practical seated posture.

Image 2 role

Supply shallow water, luminous reeds, wooden survey platform, violet fog, reflected green light, and layered marsh depth.

Prompt goal

Create an original field cartographer with one simple blank map and natural hands while the environment remains secondary.

Generated result

The output builds a new working scene where platform perspective and reflected light support the character's clear role.

Review checklist

Face and build; Two hands and one map; Reflections and platform contact

Common failure

Maps invite nonsensical text, fingers can merge with paper, and reflected light can tint the face enough to change identity.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration; any map markings are decorative and not factual, branded, or franchise-derived. Provenance: fantasy-marsh-cartographer-triptych.

Image 1 role

Anchor the fictional adult's narrow face, shaved hair, brown skin, athletic build, expression, and forward walking posture.

Image 2 role

Guide eroded sandstone arches, bright side sun, dust haze, indigo cloth accents, and a long open travel path.

Prompt goal

Create an original courier with face visible, two practical bags, wind-aware clothing, and no weapons or known faction symbols.

Generated result

The generated full-body scene coordinates dust, sunlight, and fabric movement while keeping the character as the only subject.

Review checklist

Face and skin tone; Walking anatomy; Bag count and original symbols

Common failure

Moving poses can distort feet and hands, duplicate bags, or turn practical clothing into ornate franchise-like armor.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original desert-world demonstration; not a named property, commission, or live model test. Provenance: fantasy-desert-courier-triptych.

Image 1 role

Preserve the fictional adult's broad face, long gray braid, dark eyes, older age cues, hands, and calm seated posture.

Image 2 role

Provide timber shelves, mountain window light, drying herbs, ceramic tools, muted earth colors, and practical work surfaces.

Prompt goal

Create an original herbalist using one bowl and a small herb bundle with no labels, magical text, or excessive decoration.

Generated result

The generated interior connects the subject's hands to a simple task and matches cool window light with warm timber surfaces.

Review checklist

Older facial identity; Hands and bowl interaction; No readable labels or unsafe claims

Common failure

Small tools can duplicate, labels can appear as invented writing, or stylization can remove the subject's age and facial detail.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration; not medical guidance, a real apothecary claim, customer work, or benchmark. Provenance: fantasy-mountain-herbalist-triptych.

Image 1 role

Preserve the fictional adult's face, straight dark hair, brown eyes, expression, shoulder line, and centered portrait framing.

Image 2 role

Provide rain-lit streets, cobalt palette, warm window accents, reflected signs without text, and cinematic depth.

Prompt goal

Create an original illustrated city portrait with one focal adult and general hand-painted qualities rather than named-style imitation.

Generated result

The generated image coordinates rain, clothing, and background light while keeping facial identity available for review.

Review checklist

Face and hairstyle; Rain and hand consistency; No recognizable franchise cues

Common failure

Strong stylization can flatten facial distinctions, invent text-like marks, or introduce inconsistent rain, hands, and clothing seams.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned synthetic demonstration generated with OpenAI; not a customer result, franchise image, or cross-model benchmark. Provenance: anime-city-triptych.

Image 1 role

Anchor the fictional adult's oval face, wavy brown hair, green eyes, expression, simple shirt, and three-quarter pose.

Image 2 role

Guide sloping streets, coastal roofs, blue-hour sky, warm windows, sea haze, and a wider environmental composition.

Prompt goal

Create an original hand-painted scene with the adult readable in the foreground and no named studio, artist, or character reference.

Generated result

The output combines facial and hair cues with a new coastal setting and matches evening color across the whole composition.

Review checklist

Facial identity; Perspective and body anatomy; Original buildings and symbols

Common failure

A wider scene can shrink facial detail, bend street perspective, or turn general visual cues into familiar property-like motifs.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned generated demonstration; not a user commission, customer result, franchise image, or repeatability proof. Provenance: anime-coastal-triptych.

Image 1 role

Preserve the fictional adult's square face, short auburn hair, glasses, expression, body scale, and relaxed standing pose.

Image 2 role

Provide an empty rural platform, sunrise haze, long tracks, amber edge light, utility poles, and open sky.

Prompt goal

Create one original illustrated traveler with a readable face, safe distance from the track, and no train logo or written signs.

Generated result

The generated scene uses platform lines to frame the adult and matches sunrise light across hair, clothing, and ground.

Review checklist

Glasses and face; Track perspective; Safe pose and original signage

Common failure

Glasses can distort, rail lines can converge incorrectly, or the model can add unreadable signs and branded train details.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original setting; not a studio imitation, franchise scene, customer work, or benchmark. Provenance: anime-train-platform-triptych.

Image 1 role

Anchor the fictional adult's long black hair, round face, dark eyes, expression, cardigan, and natural seated proportions.

Image 2 role

Guide tall wood shelves, window dust, green reading lamp, afternoon light, desk geometry, and warm muted colors.

Prompt goal

Create an original hand-painted reader with two natural hands, one closed book, and no readable titles or named-style request.

Generated result

The output integrates soft window light and shelf depth around a newly composed seated portrait with one clear focal subject.

Review checklist

Face and hair identity; Hands and book; No readable or branded text

Common failure

Books encourage nonsense lettering, fingers can merge with covers, or repeated shelves can create broken architectural lines.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult demonstration using general animation-inspired qualities; not a named studio or artist imitation. Provenance: anime-library-triptych.

Image 1 role

Preserve the fictional adult's tan skin, dark curls, brown eyes, expression, body proportions, and visible lower face.

Image 2 role

Provide steep roofs, mountain depth, soft snowfall, blue reflected light, warm windows, and a narrow village path.

Prompt goal

Create an original winter traveler with face unobstructed, practical coat layers, natural hands, and no recognizable property cues.

Generated result

The generated scene matches snow light across the subject and establishes a new mountain composition with clear foreground depth.

Review checklist

Skin tone and curls; Hands and winter layers; Snow scale and perspective

Common failure

Scarves can hide identity, snow can merge with hair, and clothing details can drift toward familiar character costumes.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original winter setting; not a franchise character, customer result, or model benchmark. Provenance: anime-snow-village-triptych.

Image 1 role

Anchor the fictional adult's heart-shaped face, shoulder-length silver hair, gray eyes, expression, height, and simple walking pose.

Image 2 role

Guide paper lantern shapes without text, garden paths, violet dusk, warm pools of light, and flowering tree silhouettes.

Prompt goal

Create one original festival visitor with a clear face, understated clothing, no crowd duplicates, and no named cultural property.

Generated result

The output coordinates warm lantern light with cool dusk and builds a new garden scene around one readable adult subject.

Review checklist

Face and silver hair; Hands and walking pose; Lanterns contain no text

Common failure

Lanterns can gain fake writing, background people can duplicate, or warm light can change eye and hair color excessively.

OpenAI image generation. The provider is recorded in the asset register; the exact model version was not retained as benchmark evidence. Project-owned fictional adult and original festival setting; not a studio imitation, franchise image, or customer commission. Provenance: anime-garden-festival-triptych.