Nano Banana vs GPT Image 1.5

Head-to-head across 17 challenges

Nano Banana

33.3%

win rate

Ties

23.3%

GPT Image 1.5

43.3%

win rate

33.3% 23.3% ties 43.3%

Challenge Results

Fantasy Warrior

Text-to-Image

“Close portrait of a battle-worn paladin in ornate engraved plate armor, hair braided with small beads, faint scars and dirt on the skin, warm torchlight reflecting off metal, shallow depth of field, bokeh sparks, lifelike eyes, highly detailed texture on leather straps and cloth underlayer.”

Nano Banana
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

Nano Banana

  • + Excellent rendition of ornate engraved plate armor with high detail.
  • + Clear interpretation of 'battle-worn' with distinct facial scars and weathered metal.
  • + Strong composition and lighting with clear focus on the subject.
  • Beads in the hair are somewhat generic and less detailed.
  • Skin texture looks a bit more like a digital painting compared to the realism of Model B.

GPT Image 1.5

  • + Incredible lifelike skin texture with realistic dirt and fine pores.
  • + Excellent adherence to the 'hair braided with small beads' prompt with very detailed jewelry.
  • + Masterful use of warm torchlight and bokeh sparks for a cinematic feel.
  • The armor engraving is slightly less defined than in Model A.
  • Slight anatomical ambiguity where the braid meets the shoulder armor.

Verdict: Both models followed the prompt exceptionally well, but GPT Image 1.5 stands out for its superior lifelike quality, specifically in the skin texture and the intricate detailing of the beads in the hair. While Nano Banana produced excellent engraved armor and a strong character design, GPT Image 1.5 feels more like a high-end cinematic photograph with more convincing lighting and atmosphere.

Candid Street Photography

Text-to-Image

“A candid street photo of an elderly Japanese man repairing a red bicycle in light rain, reflections on wet pavement, shallow depth of field, 50mm lens, natural skin texture, imperfect framing, motion blur from passing cars, cinematic but realistic, no stylization.”

Nano Banana
GPT Image 1.5

AI Judge Analysis

Nano Banana

  • + Excellent color grading and cinematic atmosphere
  • + High detail in the background bokeh and lighting
  • + Clean bike geometry
  • Fails to include motion blur on the passing car
  • The framing is very centered and staged despite the request for 'imperfect framing'

GPT Image 1.5

  • + Perfect adherence to 'motion blur from passing cars'
  • + Achieves a more 'candid' and 'imperfectly framed' street photography look
  • + Superior skin texture and realistic rain interaction
  • The car in the background has some minor structural incoherence due to the blur
  • The bike's handlebars and basket area are slightly cluttered and messy

Verdict: GPT Image 1.5 followed the technical requirements of the prompt much better, specifically capturing the motion blur of the passing car and the requested 'imperfect framing' of a candid street shot. While Nano Banana produced a beautiful, clean image, it felt too static and staged, failing to incorporate the motion blur that was central to the prompt's atmosphere.

Man and Car in California

Editing
Edit instruction

“Make a photo of the man driving the car down the California coastline”

Source
Nano Banana
GPT Image 1.5
50% wins 0% ties 50% wins

AI Judge Analysis

Nano Banana

  • + Perfectly preserves the source car's design and perspective.
  • + Excellent motion blur on the wheels and road adds realism.
  • + Captures the iconic California coastline aesthetic perfectly.
  • The man is rendered too small and lacks detail from the source image.

GPT Image 1.5

  • + Excellent preservation of the man's facial features and hairstyle.
  • + Strong composition that places the viewer closer to the subject.
  • + Accurately represents the coastline background with palm trees.
  • The car interior/windshield has some perspective and scale issues compared to the source.
  • The man's hand on the steering wheel is poorly rendered.

Verdict: Nano Banana excels at preserving the exact car from the source image and placing it in a high-quality, dynamic scene with realistic motion blur. However, GPT Image 1.5 does a much better job of integrating the specific man from the source image, capturing his face and hair with far more detail than Nano Banana, which relegates him to a small silhouette. Nano Banana is the better choice for car realism, while GPT Image 1.5 is the winner for subject preservation.

Geometric Composition

Text-to-Image

“A glass cube on a wooden table. Inside the cube is a small blue sphere. On top of the cube sits a red book. A green plant is behind the cube, partially visible through the glass. Soft window light from the left.”

Nano Banana
GPT Image 1.5
0% wins 67% ties 33% wins

AI Judge Analysis

Nano Banana

  • + Excellent photorealistic lighting and depth of field
  • + Natural texture on the red book and wooden table
  • + Sophisticated composition with soft window light
  • The blue sphere appears to be floating mid-air, which defies gravity
  • The plant is mostly to the side rather than 'behind' the cube as requested

GPT Image 1.5

  • + Accurate placement of all objects according to the spatial prompt
  • + Clear visibility of the plant through the glass
  • + Realistic reflections on the base of the cube
  • Lighting is a bit flat compared to Model A
  • The perspective on the book's alignment with the cube surface is slightly skewed

Verdict: Nano Banana creates a more visually stunning and photorealistic image with beautiful lighting, but it fails on physics by making the sphere float. GPT Image 1.5 followed the spatial instructions more literally, placing the plant directly behind the cube and seating the sphere firmly on the base, making it the better choice for prompt adherence.

Modern Clean Menu

Text-to-Image

“Modern minimalist restaurant menu design, white background with colorful food photos in grid, sections for appetizers/pizza/mains, bold sans-serif fonts, vibrant accents, clean professional layout for casual dining.”

Nano Banana
GPT Image 1.5
0% wins 33% ties 67% wins

AI Judge Analysis

Nano Banana

  • + Perfect adherence to the grid-style photo request.
  • + Clean, minimalist aesthetic with professional bold typography.
  • + High-quality, vibrant food photography with a consistent yellow border.
  • Several typos in section headers (APPETIERS) and menu items (Margheiita).
  • The pricing is unrealistic/repetitive for the variety of items shown.

GPT Image 1.5

  • + Excellent text rendering with almost no spelling errors.
  • + Sophisticated layout with descriptions and more realistic pricing.
  • + Logical pairing of food photos alongside their corresponding sections.
  • Does not strictly follow the 'grid' request for food photos as requested in the prompt.
  • The layout is slightly more traditional than the modern minimalist grid style requested.

Verdict: Nano Banana followed the specific layout instructions much better, providing a clear grid of photos and bold sans-serif headers, though it suffered from spelling errors. GPT Image 1.5 produced a more functional and legible menu with better text quality, but it failed to implement the requested grid arrangement for the images. Nano Banana is the likely winner for better adherence to the visual 'grid' and 'minimalist' design brief.

Bald man challenge

Editing
Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After
Nano Banana
Before After
GPT Image 1.5
100% wins 0% ties 0% wins

AI Judge Analysis

Nano Banana

  • + Excellent preservation of the original face and skin texture
  • + Hair integrates well with the existing beard and sideburns
  • + Natural-looking flow and lighting on the hair
  • The glasses frames are slightly warped/thinned where the hair overlaps

GPT Image 1.5

  • + Good hair texture and density
  • + Perfectly preserves the background and clothing
  • Significantly alters the person's facial features and bone structure
  • Creates a 'double' eyebrow/forehead wrinkle effect
  • The head shape looks unnaturally tall or distorted at the top

Verdict: Nano Banana is the clear winner because it successfully added the requested hair while keeping the original person's face identical to the source image. In contrast, GPT Image 1.5 modified the facial features so much that the person is no longer recognizable, and it introduced strange artifacts around the forehead and eyebrows.

Night Sky Transformation

Editing
Edit instruction

“Change the scene to night: a deep, dark sky with subtle, glistening stars visible behind the mountain.”

Before After
Nano Banana
Before After
GPT Image 1.5
50% wins 50% ties 0% wins

AI Judge Analysis

Nano Banana

  • + Perfect source preservation of the village and terrain details.
  • + Realistic transitions between the dark sky and the mountain peaks.
  • + Maintains the warm glow of the village lights from the original image.
  • The stars are somewhat uniform and lack the complexity of a true night sky.

GPT Image 1.5

  • + Very beautiful and detailed starry sky with nebulae-like clusters.
  • + Effective deep, dark night atmosphere.
  • Failed to preserve the source image; it significantly darkened and altered the village lights.
  • The mountain silhouette was slightly altered compared to the original composition.

Verdict: Nano Banana is the winner because it successfully executed the edit instruction while perfectly preserving the source image's village and terrain details. GPT Image 1.5 created a more visually stunning sky, but it essentially regenerated the bottom half of the image, losing much of the original character and detail of the lit town.

Over-the-top cartoon caricature

Editing
Edit instruction

“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”

Source
Nano Banana
GPT Image 1.5
100% wins 0% ties 0% wins

AI Judge Analysis

Nano Banana

  • + Captures the subject's likeness effectively within a stylized caricature format.
  • + Cleverly integrates all three themes: TV news (desk/mic), hockey (background/jersey/helmet), and dogs.
  • + Features clean, legible text and humorous details like the 'Breaking: Dog scores goal!' screen.
  • The muscular arm feels somewhat disconnected from the specific 'hockey' request compared to more obvious gear.
  • Character's hair color is a bit more reddish-brown than the source image.

GPT Image 1.5

  • + Excellent adherence to the 'exaggerated and humorous' instruction with the wide, expressive smile.
  • + Strong preservation of facial features and hair color from the source image.
  • + Very high visual quality with vibrant colors and dynamic lighting.
  • The hand holding the microphone is poorly rendered with an incorrect finger count/structure.
  • The dog on the right wearing the helmet has a slightly warped facial structure.

Verdict: Nano Banana creates a more cohesive storytelling scene with witty details and clean lines, though it shifts the subject's hair color. GPT Image 1.5 does a better job of preserving the subject's specific facial likeness and hair, but suffers from significant anatomical errors in the hand holding the microphone. Nano Banana is the winner for its superior composition, legible humorous text, and higher overall technical consistency.

Adorable Baby Animals in Sunny Meadow

Text-to-Image

“Hyper-photorealistic scene of fluffy baby animals—a golden retriever puppy, tabby kitten, baby bunny, and red fox kit—with big expressive eyes and ultra-detailed soft fur, playfully chasing butterflies and tumbling together in a lush wildflower meadow, warm golden sunrise light with god rays and dew sparkles, joyful wholesome vibe, 8K masterpiece.”

Nano Banana
GPT Image 1.5
100% wins 0% ties 0% wins

AI Judge Analysis

Nano Banana

  • + Perfect inclusion of all four requested species (puppy, kitten, bunny, fox).
  • + Clear and stylized 'god rays' lighting from the sunrise.
  • + High clarity and cleanliness in the illustration style.
  • The 'fox' resembles a corgi more than a fox kit.
  • Feels more like a digital painting than the requested 'hyper-photorealistic' scene.

GPT Image 1.5

  • + Captured the 'tumbling' and playful energy accurately.
  • + Realistic and beautiful lighting with authentic dew sparkles.
  • + High-quality fur texture and expressive facial features.
  • Anatomy errors: The kitten has too many toes and paws pointing in strange directions.
  • The bunny is partially merged with the fox and the kitten.

Verdict: Nano Banana followed the prompt by including all animals as distinct figures and achieved a very clean composition, though the fox resembles a dog and the style is quite 'AI-generic'. GPT Image 1.5 achieved much better realism, lighting, and playful energy, but it suffers from significant anatomical glitcing, especially with the kitten's paws. Nano Banana is the winner for providing a more coherent and anatomically correct image with all requested elements clearly visible.

Victorian Greenhouse Oasis

Text-to-Image

“Hyper-photorealistic interior of a lush Victorian glass greenhouse filled with exotic tropical plants, vibrant blooming orchids, tall ferns, colorful butterflies in flight, sunlight filtering through ornate glass roof creating realistic caustics and dew on leaves, intricate iron framework visible, misty atmosphere, 8K masterpiece.”

Nano Banana
GPT Image 1.5

AI Judge Analysis

Nano Banana

  • + Incredible level of detail in the intricate iron scrollwork of the Victorian structure.
  • + Features a wide variety of plant species and orchid colors as requested.
  • + The composition feels expansive and creates a strong sense of place with the pathway and bench.
  • The lighting is somewhat flat and lacks the dramatic 'caustics' requested in the prompt.
  • The butterflies are small and less integrated into the lighting of the scene.

GPT Image 1.5

  • + Outstanding lighting with visible god-rays, water droplets, and realistic dew on the foliage.
  • + Excellent adherence to the 'misty atmosphere' and 'caustics' requirements of the prompt.
  • + The butterflies are well-rendered and dynamic within the composition.
  • The ironwork structure is slightly less intricate and detailed compared to Model A.
  • The scale of the orchids in the foreground feels a bit large relative to the background trees.

Verdict: Both models produced high-quality results, but GPT Image 1.5 captured the requested atmosphere much better, specifically the misty air, dew-covered leaves, and filtered sunlight. While Nano Banana excelled at the architectural complexity of the ironwork, GPT Image 1.5 felt more like a 'masterpiece' due to its superior lighting and rendering of textures like water and light rays.

Heroic Super Hero Portrait

Text-to-Image

“Hyper-photorealistic full-body portrait of a female superhero standing triumphantly on a New York skyscraper rooftop at golden sunset, wearing a classic modest superhero costume with flowing cape, chest emblem, gloves, and boots in red and blue colors, practical design, short hair, strong determined heroic expression looking into the distance, powerful confident stance with hands on hips and cape billowing dramatically in the wind, detailed urban cityscape background, warm natural sunlight with sharp shadows and fabric highlights, ultra-sharp textures on suit, hair, and concrete, 8K masterpiece, empowering family-friendly style.”

Nano Banana
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

Nano Banana

  • + Excellent adherence to the 'modest' attire prompt with a full bodysuit.
  • + Superior lighting and atmospheric integration with the sunset.
  • + Very clean finger and hand rendering on the hips.
  • The city background is somewhat blurry compared to the subject.
  • The cape physics are a bit stiff on the left side.

GPT Image 1.5

  • + Highly detailed and sharp city background featuring recognizable landmarks.
  • + Strong textures on the costume and leather boots.
  • + Good interpretation of 'short hair' with a more distinct pixie cut.
  • Failed the 'modest' requirement by generating a short skirt.
  • The cape appears to be floating unnaturally behind her without a clear attachment point at the shoulders.
  • The left hand (resting on hip) has anatomical irregularities and merged fingers.

Verdict: Nano Banana followed the prompt much more accurately, specifically adhering to the request for a 'modest' costume by providing a full-length bodysuit, whereas GPT Image 1.5 generated a short skirt. While GPT Image 1.5 had a more detailed urban background, Nano Banana produced a more coherent character with better lighting and cleaner hand details.

Neutral Expression to Genuine Smile

Editing
Edit instruction
{
  "action": "image_edit",
  "reference": "uploaded neutral portrait",
  "change": "Warm genuine Duchenne smile: lips curved up, slight natural teeth, soft eye crinkles, subtle cheek raise",
  "details": "Realistic smiling skin (dimples if present, soft cheek shadows), slightly brighter eyes; keep exact eye shape/color/iris",
  "preserve_exact": "Face identity/structure, eyes/nose/lips/eyebrows, hair, skin texture/pores/freckles, makeup, clothing, head pose, background, lighting, shadows, framing",
  "no_changes": "No face shape change, no new features, no gaze shift, no hair/clothing/lighting/background edits",
  "style": "Ultra-photorealistic 8K portrait, sharp face focus, natural soft lighting, realistic skin glow"
}
Before After
Nano Banana
Before After
GPT Image 1.5
33% wins 33% ties 33% wins

AI Judge Analysis

Nano Banana

  • + Perfect preservation of non-edited areas like hair and background.
  • + Highly realistic integration of the smile with natural skin folding around the mouth.
  • + Maintains the original eye shape while adding subtle brightness.

GPT Image 1.5

  • + Successfully creates a warm Duchenne smile.
  • + Good preservation of the subject's identity and facial features.
  • + Realistic teeth and lip curvature.
  • Noticeable blurring/loss of detail in the hair on the left side compared to the original.
  • The eyes feel slightly more squinted/altered than the original shape requested for preservation.

Verdict: Nano Banana is the winner because it successfully applied the smile edit while keeping the rest of the image (hair, background, skin texture) exactly the same as the source. GPT Image 1.5 also produced a good smile, but it introduced unnecessary blurring and changes to the hair texture on the left side of the image.

Intricate Floral Mandala

Text-to-Image

“Perfectly symmetrical mandala made entirely of real flowers, petals, leaves, fruits, and seeds in vibrant natural colors, intricate layered patterns with radial symmetry, top-down view on a soft neutral background, hyper-detailed organic textures and subtle shadows, photorealistic, 8K masterpiece.”

Nano Banana
GPT Image 1.5
0% wins 100% ties 0% wins

AI Judge Analysis

Nano Banana

  • + Excellent photorealism with distinct textures for orange slices, chia seeds, and rose petals
  • + Maintains strong radial symmetry with a very organic, handmade feel
  • + Subtle, realistic shadows provide a sense of depth and physical placement on the canvas

GPT Image 1.5

  • + Exceptional, near-perfect geometric symmetry
  • + Very dense and intricate layering of leaves and flowers
  • + Bright, vibrant color palette that creates a striking visual impact
  • Textures feel slightly more illustrative or 'rendered' compared to the high realism of Model A
  • The shadows are very uniform, making the elements look a bit more flat or digitally placed

Verdict: Nano Banana wins this challenge by delivering a more 'real' result that perfectly balances the organic textures of fruits and seeds with the requested symmetry. While GPT Image 1.5 has more precise geometric alignment, it leans toward a digital illustration style, whereas Nano Banana feels like a high-resolution 8K photograph of actual objects placed on a textured surface.

Studio Ghibli Anime Style

Editing
Edit instruction

“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”

Source
Nano Banana
GPT Image 1.5
50% wins 0% ties 50% wins

AI Judge Analysis

Nano Banana

  • + Perfect source preservation of the original poses and expressions.
  • Completely failed the stylistic edit request.
  • The image remains a photograph with almost no discernible Ghibli or illustrative qualities.

GPT Image 1.5

  • + Excellent adherence to the Studio Ghibli art style with hand-painted textures and soft lighting.
  • + Captures the warm, nostalgic pastel color palette perfectly.
  • + Faithfully recreates the iconic poses and composition of the source meme in the new style.
  • The character in the red dress has slightly different hair volume compared to the source.

Verdict: Nano Banana failed to perform the edit, returning what is essentially the original photograph with minor lighting tweaks. GPT Image 1.5 successfully transformed the entire scene into a high-quality Studio Ghibli illustration, maintaining the recognizable composition and expressions of the 'Distracted Boyfriend' meme while perfectly applying the requested artistic style.

Golden Hour Stroll

Editing
Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

Source
Nano Banana
GPT Image 1.5
67% wins 0% ties 33% wins

AI Judge Analysis

Nano Banana

  • + Successfully added hair blowing in the wind effect while maintaining facial likeness.
  • + Added a natural amount of flying leaves with varied colors.
  • + Maintained high source preservation, keeping the background and character very consistent.
  • Most leaves have a slightly 'stamped on' look without much motion blur.

GPT Image 1.5

  • + Excellent hair motion that flows more naturally across both sides of the head.
  • + The leaves have consistent color and lighting that matches the surrounding trees.
  • + Very high preservation of the original image details and character features.
  • The large volume of leaves arguably makes the image feel cluttered rather than just 'lively'.
  • A few leaves appear to float awkwardly in front of the dog's face.

Verdict: Both models performed excellently at this editing task, maintaining near-perfect source preservation while fulfilling the prompt instructions. Nano Banana creates a cleaner look with fewer, more autumnal leaves, but GPT Image 1.5 is the winner because the hair flow is more realistic and the color of the leaves is better integrated with the existing environment.

Vintage Cafe Logo

Text-to-Image

“Vintage minimalist restaurant logo for "Caffè Florian", retro cloche dome with steam and "Est. 1720" banner, classic typography, warm brown and cream tones, subtle texture on light background, vector emblem style.”

Nano Banana
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

Nano Banana

  • + Perfectly follows the color scheme and light background request.
  • + Clean, professional vector emblem style with accurate text rendering.
  • + Excellent composition that feels like a real minimalist restaurant logo.
  • The steam is a bit stylized and merged into the dome design rather than rising from it.

GPT Image 1.5

  • + Great use of texture and shading on the cloche dome.
  • + Typography is elegant and fits the vintage aesthetic well.
  • + Clearly identifiable steam rising from the cloche.
  • Completely ignored the request for a light background.
  • Less minimalist than Model A, moving more toward an illustration than a vector logo.

Verdict: Nano Banana adhered much better to the prompt's technical requirements, specifically the light background and minimalist vector style. While GPT Image 1.5 has beautiful texture and typography, it failed a key part of the visual instruction (background color) and feels less like a functional logo than Nano Banana.

Apollo 11: Journey to Tranquility

Text-to-Image

“Create a clean, modern vector infographic poster about the Apollo 11 mission. NASA-inspired palette (navy, white, muted red, light gray). Flat-vector style, crisp lines, consistent iconography, subtle gradients only. Steps (stop at landing): 1. Launch (Saturn Vicon) 2. Earth Orbit (Earth + orbit ring icon) 3. Translunar (trajectory arc icon) 4. Lunar Orbit (Moon + orbit ring icon) 5. Descent (lunar module descending icon) 6. Landing (lunar module on the surface icon) Small supporting elements (minimal text): • Crew strip: three silhouette icons with only last names: Armstrong, Aldrin, Collins. • Landing site marker: Moon pin labeled "Tranquility" only. Layout constraints: generous margins, large readable labels, clean background with subtle stars. Vector-only, print-poster look, high resolution.”

Nano Banana
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

Nano Banana

  • + Excellent text rendering and readability.
  • + Clean, consistent icon set with a professional flat-vector aesthetic.
  • + Includes all requested steps in a logical horizontal flow.
  • The 'Descent' and 'Lunar Orbit' icons are identical.
  • The lunar module in the 'Landing' section has a slightly surreal, nonsensical structure.

GPT Image 1.5

  • + Strong visual hierarchy with a vertical progression that feels like a poster.
  • + Followed the specific icon requests for each step more accurately (e.g., Earth vs Moon distinction).
  • + Adheres very well to the NASA-inspired color palette.
  • Text is cropped at the top and bottom of the frame.
  • The iconography for 'Launch' incorrectly shows the rocket alongside Earth instead of just the rocket icon.
  • Icons and text are crowded towards the edges.

Verdict: Nano Banana produces a much cleaner and more professional infographic with nearly perfect text rendering, though it repeats icons for two different steps. GPT Image 1.5 has a better compositional flow for a poster and follows the icon descriptions more closely, but the image is poorly framed and cropped. Nano Banana is the better choice for a usable finished product due to its clarity and consistent style.

Nano Banana

Gemini 2.5 Flash Image is optimized for image understanding and generation, offering a balance of price and performance with fast and efficient image generation and editing capabilities.

GPT Image 1.5

OpenAI's state-of-the-art image generation model with better instruction following and adherence to prompts