What kind of AI image prompt is this?

This is a Nano Banana 2 AI image prompt for Ultra-Photorealistic Selfie Collage of a Woman in a Red Dress. It is designed as a ready-to-copy example so you can study the wording, match the visual style, and regenerate a similar result. The example focuses on Illustration & 3D, Photography / Poster Design direction. Keep the composition cues that matter most—subject, lighting, materials, and camera language—then swap only the parts that should change for your own brief.

How do I use this prompt?

Copy the prompt, open Wenimg with the matching model, and generate once before making edits. Start from the exact wording so you can judge what the example is actually doing.

What is this prompt best for?

It works best for Illustration & 3D, Photography / Poster Design visuals where you want a clear style reference instead of writing every art-direction detail from scratch. Use it when you need a fast starting point for campaign drafts, concept exploration, or style matching. If the output drifts, keep the original lighting and composition language and change only the subject nouns.

Wenimg

Generate new or upload & edit...

Ultra-Photorealistic Selfie Collage of a Woman in a Red Dress

Prompt

Ultra-photorealistic vertical Instagram Story-style 3x2 selfie collage featuring six candid frames of the same beautiful young woman, with strict face and identity consistency across every panel. The mood is flirty, confident, feminine, slightly seductive, and effortlessly glamorous, like a cozy night-in or date-night Instagram Story. She has soft Mediterranean beauty, warm glowing skin, realistic pore texture, natural facial asymmetry, fluffy dark brows, almond-shaped brown eyes, long lashes, and full glossy nude-pink lips with a soft wet shine. Her makeup is elegant and flattering: luminous skin, soft blush, subtle bronzer, lightly defined eyes, and glossy lips. She is wearing a flirty fitted red dress with thin straps and a feminine, slightly romantic texture. Her hair is long, dark brown, softly tousled, and voluminous, styled in a relaxed glamorous way. Add delicate layered gold necklaces, small gold earrings, a gold bracelet, rings, and dark red or burgundy manicured nails. Create a playful 3x2 selfie collage with six different poses and expressions of the same woman: Top left: close selfie with a soft pout Top right: soft smiling selfie Middle left: playful wink selfie Middle right: side-angle selfie holding a glass of red wine Bottom left: close-up beauty selfie with the wine glass slightly visible in the foreground Bottom right: blowing-a-kiss pose with one hand near her lips, holding or posing beside a glass of red wine All six panels should take place in the same cozy indoor apartment at night, with a warm intimate atmosphere. The background should feel elegant and lived-in, with soft warm ambient lighting, a sofa or bed, neutral decor, subtle candles or decor objects, and dark windows suggesting nighttime. Keep the environment clean, realistic, and stylish. Add a subtle Instagram Story interface overlay across the top, including a profile circle, the username “ozgedoner”, the time “2s”, a progress bar, three dots, and a close icon. The interface should look minimal, clean, and realistic. The camera style should feel like real iPhone front-camera selfies: casual, flattering, slightly candid, sharp but natural, not studio-lit. Use warm indoor lighting, soft shadows, realistic skin texture, glossy lip reflections, visible fabric texture, natural hand anatomy, and a soft luxury apartment vibe. The final result should feel like a real flirty Instagram Story collage from the same night. Negative Prompt low quality, blurry face, pixelated, cartoon, illustration, 3D render, plastic skin, AI doll face, over-smoothed skin, inconsistent face, different woman in each panel, distorted lips, overfilled lips, bad hands, extra fingers, fused fingers, broken nails, warped wine glass, messy collage layout, bad anatomy, wrong username, messy text, watermark, harsh flash, overexposed skin, fake fabric texture, stiff pose, unnatural expressions Nano Banana 2 @astronomerozge1 🍷 Prompt: Ultra-photorealistic vertical Instagram Story-style 3x2 selfie collage featuring six candid frames of the same beautiful young woman, with strict face and identity consistency across every panel. The mood is flirty, confident, feminine, slightly seductive, and effortlessly glamorous, like a cozy night-in or date-night Instagram Story. She has soft Mediterranean beauty, warm glowing skin, realistic pore texture, natural facial asymmetry, fluffy dark brows, almond-shaped brown eyes, long lashes, and full glossy nude-pink lips with a soft wet shine. Her makeup is elegant and flattering: luminous skin, soft blush, subtle bronzer, lightly defined eyes, and glossy lips. She is wearing a flirty fitted red dress with thin straps and a feminine, slightly romantic texture. Her hair is long, dark brown, softly tousled, and voluminous, styled in a relaxed glamorous way. Add delicate layered gold necklaces, small gold earrings, a gold bracelet, rings, and dark red or burgundy manicured nails. Create a playful 3x2 selfie collage with six different poses and expressions of the same woman: Top left: close selfie with a soft pout Top right: soft smiling selfie Middle left: playful wink selfie Middle right: side-angle selfie holding a glass of red wine Bottom left: close-up beauty selfie with the wine glass slightly visible in the foreground Bottom right: blowing-a-kiss pose with one hand near her lips, holding or posing beside a glass of red wine All six panels should take place in the same cozy indoor apartment at night, with a warm intimate atmosphere. The background should feel elegant and lived-in, with soft warm ambient lighting, a sofa or bed, neutral decor, subtle candles or decor objects, and dark windows suggesting nighttime. Keep the environment clean, realistic, and stylish. Add a subtle Instagram Story interface overlay across the top, including a profile circle, the username “ozgedoner”, the time “2s”, a progress bar, three dots, and a close icon. The interface should look minimal, clean, and realistic. The camera style should feel like real iPhone front-camera selfies: casual, flattering, slightly candid, sharp but natural, not studio-lit. Use warm indoor lighting, soft shadows, realistic skin texture, glossy lip reflections, visible fabric texture, natural hand anatomy, and a soft luxury apartment vibe. The final result should feel like a real flirty Instagram Story collage from the same night. Negative Prompt low quality, blurry face, pixelated, cartoon, illustration, 3D render, plastic skin, AI doll face, over-smoothed skin, inconsistent face, different woman in each panel, distorted lips, overfilled lips, bad hands, extra fingers, fused fingers, broken nails, warped wine glass, messy collage layout, bad anatomy, wrong username, messy text, watermark, harsh flash, overexposed skin, fake fabric texture, stiff pose, unnatural expressions Özge Döner (@astronomerozge1) Nano Banana 2 @astronomerozge1 💛 Prompt: Ultra-photorealistic vertical Instagram Story style 3x2 selfie collage, six candid frames of the same beautiful young woman, consistent face and identity across every panel. Soft Mediterranean beauty, warm glowing skin, real pore texture, natural asymmetry, fluffy dark brows, almond eyes, plump glossy pink lips with wet high-shine gloss, soft nude-pink makeup, dewy cheeks, subtle bronzer. She wears a pale butter-yellow halter-neck dress with a deep elegant neckline, soft draped fabric, feminine summer look. Hair is dark brown, messy chic low bun with loose face-framing strands, effortless and slightly undone. Small gold hoop earrings, delicate rings, glossy pale-pink manicured nails. Create a playful Instagram Story selfie grid: top row two close selfies with soft pout, middle row one wink selfie and one side-angle kissy pose, bottom row one extreme lip-gloss close-up and one blowing-a-kiss pose with hand near lips. Same indoor apartment background in every frame, soft daylight, white walls, doorway, black hanging lamp, natural home atmosphere. Add subtle Instagram story UI at the top: profile circle, username “ozgedoner”, time “2s”, progress bar, three dots and close icon, clean and realistic. Camera style: iPhone front-camera realism, slightly candid, sharp but natural, no studio look. Warm daylight, soft shadows, realistic skin, glossy lip reflections, fabric texture, natural hand anatomy. Pinterest-core it-girl selfie aesthetic, playful, pretty, confident, effortless. Negative prompt: low quality, blurry face, pixelated, cartoon, illustration, 3D render, plastic skin, AI doll face, inconsistent face, different woman in each panel, distorted lips, overfilled lips, bad hands, extra fingers, fused fingers, broken nails, distorted sunglasses, messy text, wrong username, watermark, harsh studio lighting, overexposed skin, fake fabric, unnatural pose. —

Additional details

Stats: 0

Model: Nano Banana 2
Date: May 28, 2026
Stats: 053
Author: Özge Döner
@astronomerozge1
View on
Categories: Illustration & 3D Photography Poster Design
Source: Open original

A practical reading of this AI image prompt, with the parts worth preserving and the parts that are safe to change.

How to reuse this AI image prompt

Prompt analysis

This page is built around one reusable image prompt, not a generic model description. The useful unit is the relationship between the subject, the visual treatment, the camera or layout instruction, and the finishing constraints. Reading those layers separately makes it easier to reproduce the reference image without blindly copying every noun. The selected model is Nano Banana 2, while the primary direction is Illustration & 3D. Together they define the search intent of this page: a Ultra-Photorealistic Selfie Collage of a Woman in a Red Dress AI image prompt that can be studied, copied, and adapted.

The prompt text below is the source material for the analysis. Its format is natural-language prose. This version is closer to natural language, so its layers must be identified from nouns, actions, camera cues, lighting, and finish phrases. When a prompt is written as a structured object, the field names are part of the teaching value: they show whether the author is controlling the subject, pose, environment, lighting, color, composition, or technical finish. When it is written as prose, sentence order becomes the main structure. In either case, the strongest instructions are the ones that describe a visible decision rather than a vague quality word such as beautiful or professional.

Start by separating identity from treatment. Identity includes the person, product, object, place, or action that must remain recognizable. Treatment includes the medium, lens feeling, lighting quality, palette, texture, atmosphere, and level of realism. Composition connects the two: it tells the model what should be prominent, where it should sit in the frame, and what can stay quiet in the background. This separation is especially useful when the first generation is close in style but wrong in subject, because it lets you replace the identity layer while keeping the visual grammar intact.

The example should therefore be read as a small art-direction system. Preserve the instructions that explain why the image looks the way it does; treat incidental details as optional until the core result is stable. If a detail is repeated, capitalized, placed in a JSON key, or described with a concrete camera or material cue, it is probably a high-priority constraint. If it appears only once as decoration, it is usually a candidate for later experimentation. That distinction gives this prompt value beyond the single image shown on the page.

The original prompt is included here so the explanation can be checked against the exact wording: Ultra-photorealistic vertical Instagram Story-style 3x2 selfie collage featuring six candid frames of the same beautiful young woman, with strict face and identity consistency across every panel. The mood is flirty, confident, feminine, slightly seductive, and effortlessly glamorous, like a cozy night-in or date-night Instagram Story. She has soft Mediterranean beauty, warm glowing skin, realistic pore texture, natural facial asymmetry, fluffy dark brows, almond-shaped brown eyes, long lashes, and full glossy nude-pink lips with a soft wet shine. Her makeup is elegant and flattering: luminous skin, soft blush, subtle bronzer, lightly defined eyes, and glossy lips. She is wearing a flirty fitted red dress with thin straps and a feminine, slightly romantic texture. Her hair is long, dark brown, softly tousled, and voluminous, styled in a relaxed glamorous way. Add delicate layered gold necklaces, small gold earrings, a gold bracelet, rings, and dark red or burgundy manicured nails. Create a playful 3x2 selfie collage with six different poses and expressions of the same woman: Top left: close selfie with a soft pout Top right: soft smiling selfie Middle left: playful wink selfie Middle right: side-angle selfie holding a glass of red wine Bottom left: close-up beauty selfie with the wine glass slightly visible in the foreground Bottom right: blowing-a-kiss pose with one hand near h

Suitable use cases

The strongest fit is Illustration & 3D work that needs a clear visual starting point: Illustration & 3D images, concept drafts, campaign visuals, products, or brand directions. Use the prompt when the brief has a defined mood or reference image but the art direction is not yet organized into model-ready language. It can support an early concept, a campaign direction, a social post, a product study, or a personal visual experiment, provided the requested subject remains within the model’s capabilities and the final output is reviewed by a person.

For a professional workflow, treat the prompt as a first-pass specification rather than a final guarantee. It is useful for comparing how Nano Banana 2 interprets lighting, materials, framing, and visual hierarchy. Generate the original once, then create a controlled variant with only one major change. That comparison tells you whether the result is being driven by the subject, the style, or the camera language. The method is more reliable than changing five clauses at the same time and guessing which edit mattered.

This structure is also helpful for learning prompt writing. A new user can identify the nouns that establish the scene, the adjectives that establish the look, and the technical phrases that limit the output. An experienced creator can use the same page as a compact reference when moving between models. The goal is not to force every image into one formula; it is to make the decisions visible enough that the creator can choose which decisions to keep, weaken, or replace.

Use the page responsibly when the prompt includes a person, a recognizable brand, a source image, or a real-world setting. Keep permissions, privacy, attribution, and platform rules in mind. The source and attribution links on this page are part of the case context, not proof that every future use is authorized. A good adaptation changes the creative brief for a legitimate purpose and checks the result before publishing or using it commercially.

How to modify it

The safest edit order is subject first, then composition, then style, then finish. Begin with the subject noun and the action or product requirement. Keep the original framing and lighting for the first variant so you can see the effect of that one replacement. Next change the composition if the image needs a different crop, camera distance, angle, or subject position. Only after those choices are stable should you tune palette, texture, atmosphere, typography, or small props.

When adapting this prompt to a new brief, keep a small preserve / replace / remove list. Preserve the cues that carry the page’s core visual identity, such as a soft directional light, a reflective material, a centered subject, a shallow depth of field, or a particular poster hierarchy. Replace nouns that belong only to the original case. Remove contradictory details rather than stacking more instructions on top of them. A shorter prompt with consistent constraints is often easier for the model to follow than a long prompt with competing camera, style, and layout demands.

If the model drifts, make one diagnostic pass before rewriting everything. First ask whether the subject is too ambiguous. If yes, make its appearance, count, pose, and position concrete. If the subject is right but the mood is wrong, adjust the lighting and palette language. If the mood is right but the layout is wrong, change only the camera or composition clause. If text, logos, hands, or small product details fail, describe their role and placement clearly, then verify the output manually because generative models can still make factual or typographic mistakes.

For repeatable work, record the model, aspect ratio, reference images, and the exact edit made for each generation. That turns this case from a one-off prompt into a small experiment. Keep the original prompt untouched, create named variants, and compare them at the same output size. The result is a useful feedback loop: the page teaches the initial structure, the generated image shows what the model actually obeyed, and the next revision is based on observed behavior rather than assumptions about a model’s capabilities.

Copy the original prompt and generate one baseline result with Nano Banana 2.
Mark each clause as identity, composition, style, lighting, material, or finish.
Replace one identity or brief-specific detail while keeping the visual treatment stable.
Compare the variant with the reference, then change one additional instruction only when the difference is understood.
Save the successful wording with the model, aspect ratio, and reference-image context.

Related case chain

A useful internal path is to move from this single case to the model collection, then to the category collection, and finally to another concrete example. The model page shows the neighboring prompt vocabulary available for Nano Banana 2; the category page narrows the journey to Illustration & 3D; the gallery exposes other images that can be compared by subject, composition, and finish. This is a case chain rather than a list of unrelated links: each step answers the next question a reader normally has after studying one prompt.

When comparing neighboring cases, change one dimension at a time. Compare two prompts from the same model to understand model-specific phrasing. Compare two prompts in the same category to see which visual decisions are transferable. Then compare a close case with a contrasting one to identify which details are truly essential to the subject. The links below are intentionally descriptive so that both people and crawlers can understand the relationship without relying on a button click or an opaque identifier.

The current case remains the reference point for the chain. Return to it when a related example introduces a different lighting setup, camera distance, material, or visual density. Copy only the clause that solves the new problem, not the entire neighboring prompt. This keeps the page’s primary intent intact while giving the reader a practical route through the broader AI image prompt library.

Where does this prompt come from?

This AI image prompt is a reusable example attributed to Özge Döner.

Data from NanoBanana Trending Prompts (https://github.com/jau123/nanobanana-trending-prompts) by MeiGen.ai, licensed under CC BY 4.0.

View the original source License: CC BY 4.0