soy_aria_cruz: Seedream 4 Vs Nano Banana Kitchen Selfie

💥 Seedream 4 VS. Nano-Banana Hoy toca poner a prueba el nuevo generador Seedream 4 que promete crear imágenes en 4K 😍 La idea fue elegir diferentes retos y situaciones para ponerlo a prueba con Nano Banana 🍌 - Integrando un producto (Monster) - Cambios de ropa + de marca - Imágenes de muy cerca (piel, ojos, boca) - Selfies - Buceando debajo del agua - Expresiones faciales - Imágenes de cuerpo completo para la consistencia del rostro Si se te ocurren más situaciones que no he tenido en cuenta, comenta cuáles y lo pongo a prueba 💕 💌 Y dime... Con cual te quedas de los dos?? 👀

Aria Cruz | Influencer AI

@soy_aria_cruz · Digital creator

INSTAGRAM · 2025-09-11Source

253likes

45comments

Remix This

Recreate with Create/ai Image Generator

Make your own AI viral video

Prompt

[Subject] A split-screen comparison graphic showing the same young woman rendered in two selfie-style kitchen breakfast scenes. She appears in her early 20s with a slim build, fair warm skin, large blue-green eyes, dark thick brows, and a friendly smiling expression in both panels. She has dark brown to black hair tied up in a messy bun or casual top bun, with loose front strands framing the face. She wears large round wire-frame eyeglasses and medium hoop earrings. In both images she wears a soft gray hoodie. Keep exactly one woman per panel, representing the same identity consistently.

[Scene setup] The entire image is a two-column comparison card divided by a dark teal vertical line and surrounded by a matching dark teal border. Left panel: the woman leans into the frame from a slightly high selfie angle in a bright white kitchen, smiling up toward the camera. A breakfast bowl filled with cereal or granola appears on a countertop in the lower-left foreground. The background includes white cabinetry, a bright archway or curved opening, and soft daylight. Right panel: the same woman appears in a more straightforward front-facing selfie in a warm, softly lit kitchen with white walls, a bright window, a potted trailing plant on a shelf, and a breakfast bowl on the counter behind or beside her. Keep the right panel slightly cleaner and calmer than the left.

[Environment] Both panels take place in cozy domestic kitchens during daytime. The left side should feel brighter, slightly more exaggerated, and more “camera-up-high” in perspective. The right side should feel softer, more realistic, and more naturally balanced. Both kitchens should be lived-in but tidy, with breakfast context implied by bowls and counters rather than elaborate food styling.

[Composition/Camera] Vertical 4:5 comparison graphic with two equal portrait panels side by side. Left panel uses a stronger downward selfie angle, creating a more distorted playful perspective with the subject leaning over the counter. Right panel uses a more natural handheld selfie angle closer to eye level. Both panels crop the subject from mid-torso upward. Preserve the split-screen layout with rounded card corners and the central divider.

[Lighting] Both images are softly daylight-lit, with the left side brighter and slightly more stylized in contrast, and the right side more natural and balanced. No harsh flash. Skin tones should stay flattering and bright. The left panel can feel a touch more polished or glossy, while the right panel should feel a bit more grounded and realistic.

[Text overlays] Keep the comparison graphic text exactly. Near the bottom center spanning both panels there is white text: `Prompt:` and below it `La chica se hace una selfie desayunando`. At the lower-left of the left panel there is bold yellow text `SEEDREAM 4`. At the lower-right of the right panel there is bold yellow text `NANO-BANANA`. Preserve all overlay text because it is part of the provided image and the comparison meaning depends on it.

[Color palette] Dominant colors are soft gray, white kitchen tones, light wood or beige countertop accents, warm skin, black hair, and dark teal framing. The yellow labels act as strong accent colors. Overall look should feel bright, clean, and creator-tutorial friendly.

[Style/Rendering] Hybrid social-media comparison cover: two hyper-real AI selfie outputs presented as a generator-vs-generator test card. The frame should feel educational and shareable, not like a normal single portrait. Clean domestic realism, soft beauty polish, and obvious side-by-side comparison structure are essential. Not a meme, not a screenshot collage without design, not a technical benchmark chart.

[Detail constraints] Do not remove the split-screen layout, the dark teal border/divider, the gray hoodie, glasses, or breakfast bowls. Keep exactly one woman in each panel with consistent identity. Preserve the Spanish prompt text and the yellow generator labels `SEEDREAM 4` and `NANO-BANANA`. Do not merge the panels into one scene. Do not add extra people, extra text, or extra kitchen clutter.

Negative prompt: single image only, no divider, no labels, no breakfast context, outdoor setting, bathroom mirror selfie, smartphone visible in hand, heavy flash, dark nightclub, office kitchen, children, couple scene, overdecorated counters, mugs everywhere, anime rendering, painterly split-screen, unreadable text, extra panels, warped face, inconsistent identity between panels.

Suggested parameters: aspect ratio 4:5 vertical; left panel selfie feel around 24-28mm equivalent, right panel around 28-35mm equivalent; aperture look around f/2.8 to f/4; steps 30-42; CFG/style guidance 6.5-8; sampler DPM++ 2M Karras or realistic lifestyle sampler; seed suggestion 638520194.

Delta prompt strategy:
1. If the comparison layout disappears: "two side-by-side portrait panels with dark teal divider and border, rounded corners"
2. If identity consistency is lost: "same brunette woman in both panels, gray hoodie, round glasses, hoop earrings, messy bun"
3. If the breakfast cue vanishes: "breakfast bowl on the counter in each kitchen scene"
4. If the left panel loses its perspective: "left panel high-angle selfie leaning over the kitchen counter, slightly more stylized"
5. If the right panel becomes too similar: "right panel calmer eye-level selfie in a cozy kitchen with plant and window light"
6. If text overlays vanish: "white text: Prompt: La chica se hace una selfie desayunando, yellow labels SEEDREAM 4 and NANO-BANANA"
7. If the environment drifts: "bright tidy home kitchen interiors, not bathroom, not café, not office"
8. If the graphic loses shareability: "generator-vs-generator social comparison cover, educational and visually clean"
9. If extra objects appear: "minimal counters with only bowls and light kitchen details"
10. If the lighting becomes harsh: "soft daylight in both panels, flattering, bright, no flash"

Why soy_aria_cruz's Seedream 4 Vs Nano Banana Kitchen Selfie Went Viral — and the Formula Behind It

This image works because it turns a very ordinary situation into a useful visual test. “A girl taking a breakfast selfie” is not a spectacular prompt on its own, and that is exactly why it is smart. Simple domestic prompts reveal model behavior quickly. You can see how each system handles identity, daylight realism, selfie perspective, background organization, and casual lifestyle mood without getting distracted by a complex scene.

The split layout also does a lot of heavy lifting. The dark border, central divider, and clear labels make the comparison readable in less than a second. That matters for social media. A comparison post succeeds when the audience understands the game immediately: same idea, different generator, pick your favorite. This frame explains that fast.

The best part is that the scenario feels genuinely useful for creators. Breakfast selfies, hoodie shots, kitchen backgrounds, and home-life moments are common use cases for AI influencer pages. So the image is not only comparing two models; it is comparing them inside a situation creators actually care about. That practical relevance makes the post stronger than a purely flashy benchmark would be.

Why This Comparison Format Performs

The first reason is low-concept accessibility. Everyone instantly understands the scene. The second reason is side-by-side clarity. The viewer is invited to compare style instead of decode context. The third reason is real-world usefulness. Kitchen selfies are common enough that differences between generators feel actionable, not abstract.

The image also benefits from balanced asymmetry. The left panel is slightly more exaggerated in perspective, which makes it more energetic. The right panel is calmer and more natural, which can make it feel more believable. That difference creates a genuine vote instead of two panels that feel interchangeable.

Signal	Evidence (from this image)	Mechanism	Replication Action
Instant comparison logic	Two equal panels, divider, generator labels, prompt text	Viewers know immediately how to engage and what to compare	Use explicit structural cues so the audience understands the test without reading the caption first
Useful everyday prompt	Simple breakfast selfie in a home kitchen	Relatable prompts make model differences feel practical for creators	Choose test prompts that map to actual creator use cases, not only fantasy scenarios
Identity consistency test	Same face, glasses, hoodie, and hairstyle across both sides	The audience can judge the model instead of getting lost in character drift	Lock 3-4 signature identity cues in every panel of the comparison
Subtle style contrast	Left feels more exaggerated, right feels more natural	Meaningful differences encourage comments and preferences	Present outputs where the distinction is visible but the underlying scenario is the same

Best Use Cases and Transfers

This layout is ideal for generator-vs-generator tests, prompt education posts, creator tool reviews, “which output do you trust more?” comparisons, and practical AI use-case breakdowns. It works especially well when the prompt is simple enough that stylistic differences become the main thing people notice.

Best fit: AI model battles. The labels and clean structure make the post inherently interactive.
Best fit: tool education for creators. The prompt is ordinary enough to feel directly reusable.
Best fit: consistency tests. Matching the same identity across both panels lets viewers compare model handling honestly.
Best fit: tutorial thumbnails. The comparison reads quickly and promises useful information.
Best fit: carousel opener slides. It creates a clear “here is the experiment” frame for the rest of the post.

It is less suitable for purely emotional storytelling. This format is analytical in function, even if it looks warm and friendly.

Transfer Recipes

Product integration variant. Keep: same panel structure and identity lock. Change: breakfast bowl to handheld product. Slot template: {same character}, split-screen model comparison, {simple use case}, clear generator labels, one prompt line
Mirror-selfie variant. Keep: domestic realism and comparison card design. Change: room type and camera angle. Slot template: {same woman}, side-by-side AI comparison, {selfie prompt}, labeled panels, social-friendly layout
Close-up beauty variant. Keep: two-panel educational format. Change: crop size and test focus. Slot template: {generator A vs generator B}, one consistent face, {beauty prompt}, clear prompt text, comparative visual framing

What the Aesthetic Is Doing

The image is effective because it keeps the prompt ordinary and lets the rendering differences do the talking. The left image feels a little more stylized and perspective-heavy. The right image feels more relaxed and natural. That is exactly the kind of distinction that audiences can talk about without needing technical vocabulary. It turns subjective style judgment into easy engagement.

The hoodie is also a smart wardrobe choice. It removes fashion complexity and keeps the comparison inside a casual daily-life setting. That means viewers focus on face, light, and realism rather than on fabric detail or branding. In comparison posts, simple clothing is often better because it reduces noise.

The breakfast bowls are small but important. They make the prompt feel grounded. Without them, the scene could just be “girl in kitchen.” With them, the image supports the written prompt more precisely, which makes the comparison feel fairer and more intentional.

Observed	Why it matters	How to recreate it
Dark teal border and divider	Turns two outputs into one clear shareable card	Use a strong frame color to unify the comparison layout
Prompt text across the lower middle	Anchors the test and reduces ambiguity	State the exact prompt or scenario directly on the image
Same hoodie and face cues in both panels	Keeps the test about model behavior instead of styling changes	Hold identity markers constant across every compared output
Different selfie energy per panel	Creates a real choice instead of redundant outputs	Let each model show its bias while preserving the same scene objective
Breakfast bowls in both kitchens	Makes the domestic prompt feel specific and honest	Include one small object that confirms the use case in each panel

Prompt Technique Breakdown

The key prompt lesson here is that comparison graphics need both image quality and editorial packaging. A good pair of outputs is not enough if the post does not explain itself quickly. The labels, the divider, and the prompt line are not decoration. They are the mechanism that makes the comparison readable and shareable.

Prompt chunk	What it controls	Swap ideas (EN, 2-3 options)
Use-case prompt	What practical scenario is being tested	breakfast selfie; kitchen morning selfie; casual home-life portrait
Identity anchors	Consistency across both generators	round glasses; gray hoodie; messy bun with loose strands
Panel differentiation	Why the audience has something real to compare	high-angle selfie vs eye-level selfie; stylized vs natural; brighter vs calmer
Comparison frame	How clearly the post communicates its structure	split-screen with divider; labeled side-by-side card; bordered A/B comparison
Prompt label text	Educational usefulness of the post	prompt subtitle; exact scenario line; test condition caption
Minimal domestic props	Specificity without clutter	breakfast bowl; tidy counter; simple kitchen details

The biggest drift risk is overcomplicating the prompt. If the scene becomes too elaborate, the audience stops comparing generators and starts comparing everything else. Simplicity is what makes the benchmark useful.

Execution Playbook for Iteration

Lock the structure first: two panels, divider, labels, and prompt text. Then lock the identity: glasses, hoodie, bun, same face. Then differentiate the angles and kitchen mood. That order makes sure the post stays useful as a comparison card even before the fine visual polish is perfect.

Use the one-change rule. First solve the domestic scene. Then solve identity. Then solve panel differences. Then solve typography and framing. This approach keeps the comparison honest and prevents the design layer from masking weak visual logic.

Run 1: Generate two kitchen breakfast selfie panels with the same woman in a gray hoodie and glasses.
Run 2: Push the left panel toward a higher-angle, slightly more stylized selfie and keep the right more balanced and natural.
Run 3: Add the breakfast bowls, dark teal divider, and border to turn the images into one coherent card.
Run 4: Overlay the prompt text and yellow generator labels, then refine the kitchen brightness and facial consistency.

If the panels look too similar, exaggerate the camera-angle difference before changing anything else. If they feel too different, bring back the same hoodie, face, and breakfast context. If the post looks cluttered, remove background kitchen details before touching the comparison text. The strongest version is simple, useful, and instantly arguable.