Head to head
Esc

Models · slot A

to navigate to pick

FLUX.2 [max] Black Forest Labs GPT Image 1.5 OpenAI

Settled by community votes across 20 shared challenges, with an AI judge weighing in on each.

FLUX.2 [max]

25.8 arena score

#10 of 62 in Text-to-Image

Skill signature · Text-to-Image

GPT Image 1.5

27.1 arena score

#7 of 62 in Text-to-Image

Top 3 in Image Editing
Vote tally

Where the votes landed

FLUX.2 [max]

28.6%

win rate

Ties

14.3%

GPT Image 1.5

57.1%

win rate

28.6% 14.3% ties 57.1%
Shared challenges 20

Challenge by challenge

The strongest take from each model on every shared challenge, with the AI judge's read.

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.”

FLUX.2 [max]
GPT Image 1.5
25% wins 50% ties 25% wins

AI Judge Analysis

FLUX.2 [max]

  • + Excellent cinematic lighting with realistic bokeh and shadows.
  • + Very high texture detail, particularly on the table grain and the leather-like book cover.
  • + Correct interpretation of the 'partially visible through glass' plant element.
  • The glass cube lacks a physical top panel, making the book appear to float slightly or rest only on the edges.
  • The plant in the background is significantly blurred compared to the main subject.

GPT Image 1.5

  • + Very clean and solid construction of the glass cube with clear edges and a visible top surface.
  • + Strong adherence to all prompt elements including the lighting direction and object placement.
  • + Realistic reflections on the blue sphere and the bottom of the cube.
  • The wood grain on the table is somewhat generic compared to Model A.
  • The plant lacks the depth and soft focus that would make the composition feel more professional.

Verdict: Both models followed the prompt perfectly, including the complex spatial relationships between the objects and the specific lighting direction. FLUX.2 [max] produced a more artistic, high-end photographic result with superior textures, but GPT Image 1.5 handled the physical logic of the glass cube better by ensuring the book had a clear surface to rest upon.

Man and Car in California

Editing
Edit instruction

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

Source
FLUX.2 [max]
GPT Image 1.5
33% wins 0% ties 67% wins

AI Judge Analysis

FLUX.2 [max]

  • + Excellent preservation of the car's model and specific design details.
  • + Accurately places the man inside the car while maintaining his scarf and dreadlocks.
  • + High-quality environment and lighting integration.
  • The man's expression is quite stern compared to the original smile.
  • The scaling of the man seems slightly small relative to the car's interior.

GPT Image 1.5

  • + Captures the man's facial expression and smile more accurately than Model A.
  • + Great dynamic composition showing the perspective of driving along the coast.
  • Significant loss of car identity, changing the grill and front details of the Rolls Royce.
  • Inaccurate hand placement/anatomy on the steering wheel.

Verdict: FLUX.2 [max] is the superior model for this edit because it preserves the specific identity of both source subjects: the exact Rolls Royce model and the man's distinct style. While GPT Image 1.5 captures the man's smile better, it fails to maintain the car's design and features mangled hand geometry.

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.”

FLUX.2 [max]
GPT Image 1.5
50% wins 0% ties 50% wins

AI Judge Analysis

FLUX.2 [max]

  • + Excellent adherence to the motion blur request for passing cars.
  • + Highly realistic skin textures and fine details on the hands.
  • + Accurately captures the 'imperfect framing' of a candid street photo.
  • The brake cables on the bicycle are a bit messy and structurally illogical.
  • The background cars have slightly surreal lighting/ghosting.

GPT Image 1.5

  • + Stronger atmospheric rain effects with visible droplets on the clothing and hat.
  • + Great attention to the reflections on the wet pavement.
  • + Composition feels very grounded and lifelike.
  • Missed the specific request for motion blur on the passing cars; the background car is static.
  • Minor anatomical confusion where the hands meet the bicycle chain area.

Verdict: FLUX.2 [max] followed the complex technical prompt more closely, particularly by including the requested motion blur on background traffic and the imperfect framing. While GPT Image 1.5 delivered a more atmospheric rain effect and beautiful reflections, it failed to incorporate the motion blur, resulting in a more static scene. FLUX.2's skin textures and adherence to the 50mm lens look make it the more successful interpretation of the 'candid street photo' aesthetic.

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.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent depiction of ornate engraving on the plate armor.
  • + Superior skin texture featuring realistic scars and dirt.
  • + Clean, professional lighting with a clear shallow depth of field.
  • The beads in the hair are a bit small and less distinct than in Model B.
  • The expression is a bit static compared to the intensity of the scene.

GPT Image 1.5

  • + Very lifelike and expressive eyes that draw the viewer in.
  • + Fantastic lighting with vibrant warm highlights and bokeh sparks.
  • + Strong adherence to the 'braided hair with small beads' part of the prompt.
  • The armor textures are slightly noisy and less defined than in Model A.
  • Over-saturation of skin tones makes the face look a bit orange in some areas.

Verdict: FLUX.2 [max] provides a more technically grounded and balanced image with superior metallic textures and clear engraving, while GPT Image 1.5 excels in emotional expression and atmospheric lighting. Ultimately, FLUX.2 [max] is preferred for its exceptional clarity and more realistic handling of materials and facial skin textures.

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.”

FLUX.2 [max]
GPT Image 1.5
33% wins 0% ties 67% wins

AI Judge Analysis

FLUX.2 [max]

  • + Strictly followed the request for a grid of food photos.
  • + Excellent use of vibrant secondary colors (yellow, red, green) for section headers.
  • + Clean, professional aesthetic that feels like a real modern layout.
  • The text is largely gibberish/placeholder text.
  • Some image artifacts present, like a floating hand in the slider photo.

GPT Image 1.5

  • + Perfect English text and logical menu pricing.
  • + Clear association between menu categories and the corresponding food photos.
  • + High visual quality of the food photography.
  • The food photos are in stacked blocks rather than a true grid layout as requested.
  • Design feels a bit more generic compared to the bold styling of the other model.

Verdict: While FLUX.2 [max] captures the 'grid' and 'bold vibrant accents' of the prompt much more effectively, GPT Image 1.5 produces a functional menu with perfect English text. However, as a design challenge, FLUX.2 [max] feels more closely aligned with the requested modern minimalist aesthetic and specific layout requirements.

Magic Burger Explosion: Fiery Photorealism Challenge

Text-to-Image

“Ad for 'Magic Burger'. Dynamic, exploded burger with all components (bun, patty, cheese, lettuce, tomato, sauce) suspended in mid-air. Emphasize photorealistic detail and a sense of motion. Dark, fiery background with glowing embers. Integrate text: 'MAGIC BURGER' as a prominent title, 'LIMITED TIME ONLY' as a secondary message, and '€6.99' in a starburst, all rendered with a fiery, glowing effect.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent photorealistic texture on the bun and patty
  • + Clean and readable text integration
  • + Clearer exploded view showing the separation of all components
  • The '€6.99' starburst looks like a flat graphic compared to the photorealistic burger
  • The lighting on the bottom bun feels slightly disconnected from the fiery background

GPT Image 1.5

  • + Excellent adherence to the 'fiery, glowing effect' for all text elements
  • + Dynamic composition with embers that feel more integrated into the scene
  • + Stronger overall atmosphere and color harmony
  • The burger ingredients are less distinctly 'exploded' and look somewhat cluttered
  • Slightly more artificial 'AI' look to the food textures compared to Model A

Verdict: Both models followed the prompt exceptionally well. FLUX.2 [max] produced a cleaner, more photorealistic burger with a better 'exploded' layout, while GPT Image 1.5 achieved a more cohesive 'fiery' atmosphere and more creative glowing text integration. FLUX.2 [max] is the likely winner due to the superior quality of the food photography itself, which is central to a food ad.

Chalkboard Menu

Text-to-Image

“Handwritten-style chalkboard menu in a cozy café, all text rendered in the exact same realistic chalk handwriting style with natural variations in letter size, slight slant, and chalk texture — no printed or digital fonts anywhere on the board. Title at the top in elegant cursive chalk handwriting: ‘TODAY’S SPECIALS – APRIL 30, 2026’. Below it, three menu items also in the same handwritten chalk style: ‘Truffle Mushroom Risotto – $24’, ‘Grilled Octopus with Lemon & Herbs – $28’, ‘Brown Butter Chocolate Chip Cookies – $9’. At the very bottom, smaller text in the identical handwritten chalk style (slightly smaller but still clearly legible with the same handwriting characteristics): ‘All items made fresh daily • Ask about our gluten-free options’. Warm ambient café lighting, visible chalk dust and smudges, realistic handwriting imperfections, no clean printed text anywhere.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Displays a high-quality chalk texture with realistic smudges on the board.
  • + Correctly renders all prompt text including the extended list of items.
  • + Features a clean, centered composition with warm overhead lighting.
  • Handwriting appears slightly too uniform, resembling a digital chalk font.

GPT Image 1.5

  • + Successfully captures a more organic, handwritten feel with variable letterforms.
  • + Accurately renders all requested text including pricing and the footer note.
  • + Chalkboard smearing and texture look highly authentic.
  • The layout is somewhat less balanced with a tighter crop at the top.
  • Cursive elements in the title are less elegant compared to the other model.

Verdict: Both models followed the complex text instructions perfectly, including the date and specific menu items. FLUX.2 [max] offers a more balanced composition and professional look, while GPT Image 1.5 provides a more convincing 'hand-drawn' feel with varied character widths and a more textured board surface.

Pose & Character Mashup

Editing
Edit instruction

“Use Image 1 as the exact pose reference and Image 2 as the character reference. Recreate the person/character from Image 2 in the exact dynamic pose and body position from Image 1. Keep the exact face, hair, clothing style/details, and expression from Image 2. Match the lighting and environment of Image 1. The final image must show the character from Image 2 performing the precise action/pose from Image 1 with perfect anatomy and natural integration.”

Source
FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent character face and accessory recreation from Image 2.
  • + Matches the atmospheric lighting and warm tones perfectly.
  • Fails significantly on the leg pose, resulting in a squat rather than the crossed leg position.
  • The right hand has severe anatomical issues (claw-like fingers).

GPT Image 1.5

  • + Follows the exact leg-crossing pose from Image 1 much more accurately.
  • + High fidelity in preserving the scarf, sunglasses, and facial hair of the character.
  • + Better hand anatomy compared to the other model.
  • The head angle is slightly less dynamic/tilted than requested in the reference pose.

Verdict: While both models successfully integrated the character from Image 2 into the environment of Image 1, GPT Image 1.5 is the clear winner for its superior pose adherence. It correctly captured the difficult leg-crossing position and maintained much better anatomical consistency, whereas FLUX.2 [max] failed to replicate the specific lower-body pose and produced distorted hands.

The Reversed Rodeo

Text-to-Image

“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent cinematic lighting and composition.
  • + Clean, high-resolution rendering with a surreal floating asteroid base.
  • + Distinct separation between foreground and the background galaxy.
  • Failed the specific reversal instruction; the astronaut is riding the horse.

GPT Image 1.5

  • + Dynamic sense of motion and detailed lunar environment.
  • + Highly detailed space suit and horse tack textures.
  • Failed the specific reversal instruction; the astronaut is riding the horse.
  • The scale of planets in the background feels cluttered and inconsistent.

Verdict: Both models failed the negative constraint to put the horse on top of the astronaut, showing the inherent difficulty models have with spatial reversal prompts. FLUX.2 [max] is the preferred image as it has much cleaner composition and lighting, whereas GPT Image 1.5 feels overly busy with conflicting planetary scales.

Outfit Transfer Challenge

Editing
Edit instruction

“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”

Source
FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Successfully replicates the exact outfit from Image 2 including the watch and rings.
  • + Preserves the face and unique vitiligo patterns with high accuracy.
  • + Matches the warm lighting and color temperature of the original scene.
  • The scarf pattern is slightly simplified compared to the source.
  • Minor change to the intensity of the sand on the cheek.

GPT Image 1.5

  • + Good replication of the scarf's fraying and layering.
  • + Accurately places the gold watch from the source image.
  • Fails the primary instruction by cropping out the face entirely.
  • The jeans color is significantly darker than the source outfit.
  • Lost the ring detail from the source's left hand.

Verdict: FLUX.2 [max] is the clear winner as it successfully transferred the clothing while keeping the person's identity and face visible as requested. GPT Image 1.5 failed the core prompt requirements by cropping the person's head out of the frame, which effectively avoided the person-preservation aspect of the challenge.

The Capybara Taxi Driver

Text-to-Image

“Photorealistic scene inside a yellow New York taxi at night. A capybara is driving, wearing a yellow taxi driver cap and a dark jacket. It has a calm, professional expression and both front paws on the steering wheel. In the back seat sits a human businesswoman in a coat, looking at her phone with a completely normal, bored expression (as if this is just another normal ride). Through the windows you can see the streets of Manhattan at night with blurred lights. Realistic taxi interior, photorealistic, detailed fur and fabric, 35mm lens, night lighting with reflections, shallow depth of field.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Exceptional photographic realism in lighting and textures
  • + Highly detailed taxi interior including the dashboard, ignition, and seats
  • + Realistic depiction of the passenger's face and phone
  • The capybara appears to have human hands in gloves rather than paws on the wheel
  • The capybara's anatomy is slightly distorted to fit a human seat and posture

GPT Image 1.5

  • + Excellent adherence to the 'front paws' requirement showing actual capybara anatomy
  • + The taxi driver cap is more iconic and professional as requested
  • + Great centered composition that highlights both the driver and passenger
  • Lower overall image resolution and clarity compared to Model A
  • The capybara's eyes look slightly artificial

Verdict: FLUX.2 [max] creates a more stunning, photorealistic environment with professional lighting and interior textures, but it fails the specific requirement for 'paws' by giving the animal human-like gloved hands. GPT Image 1.5 follows the anatomical prompt instructions much better, showing actual capybara paws on the wheel and a more appropriate uniform hat, making it the better interpretation of the specific scene requirements despite lower technical fidelity.

The Halloween Invitation

Text-to-Image

“Vintage gothic Halloween party invitation. Dark parchment poster, spooky border with webs and thorns, central glowing jack-o-lantern, bats, twisted trees, moody night sky. Add elegant gothic title text saying "Halloween Party Invitation", a small scroll banner saying "You are invited to a night of frights", and event details at the bottom: Date: 30.10.2026 Time: 7pm Location: The Arches, NYC Spooky but polished, cinematic lighting, square format.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Crystal clear text rendering for all details including the small banner.
  • + Clean, modern gothic layout with sharp borders.
  • The parchment effect feels like a digital overlay rather than an integrated texture.
  • The lighting on the pumpkin is a bit flat compared to the background.

GPT Image 1.5

  • + Excellent 'vintage' aesthetic with a cohesive dark parchment texture throughout.
  • + Dynamic, atmospheric lighting with a glowing moon and graveyard silhouette.
  • + Creative scroll banner and decorative flourishes that match the gothic theme.
  • Slightly less crisp text on the 'Location' line compared to Model A.
  • The thorn border is a bit repetitive in its pattern.

Verdict: Both models followed the prompt exceptionally well, producing accurate text and all requested elements. GPT Image 1.5 is the winner because its aesthetic perfectly captures the 'vintage' and 'moody' request with superior texture and atmosphere, whereas FLUX.2 [max] feels slightly more like a clean digital composite.

Bald man challenge

Image 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
FLUX.2 [max]
Before After
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [max]

  • + Excellent preservation of the original image's lighting, background, and clothing.
  • + The hair texture is high-resolution and matches the existing beard well.
  • + Maintains the exact facial geometry and expression of the source image.
  • The hair volume is arguably a bit exaggerated at the top, looking slightly like a wig.
  • The hair encroaches a bit too much on the ear/glasses contact point.

GPT Image 1.5

  • + Natural, realistic hair density and curl pattern that fits the character's aesthetic.
  • + Very clean integration of the hairline with the forehead.
  • + Almost perfect preservation of pixels outside the hair area.
  • Slightly altered the shape of the glasses frames.
  • Subtle changes to the nose and eye area making the person look like a slightly younger version of themselves.

Verdict: Both models did an exceptional job at adding realistic hair while preserving the scene. FLUX.2 [max] is preferred because it maintained the person's exact facial features and the original glasses perfectly, whereas GPT Image 1.5 subtly face-swapped a slightly different person onto the head, despite having a more natural-looking hair style.

Isometric Miniature Diorama Scenes

Text-to-Image

“Create a clear, 45° top-down isometric miniature 3D cartoon scene of Japan's signature dish: sushi, with soft refined textures, realistic PBR materials, gentle lighting, on a small raised diorama base with minimal garnish and plate. Solid light blue background. At top-center: 'JAPAN' in large bold text, 'SUSHI' below it, small flag icon. Perfectly centered, ultra-clean, high-clarity, square format.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent adherence to the 'cartoon' and 'soft textures' stylistic requirements.
  • + Clean and professional typography layout with correct flag placement.
  • + Minimalist approach matches the 'minimal garnish' instruction perfectly.
  • The diorama base is a bit plain compared to the '3D miniature' potential.

GPT Image 1.5

  • + High-quality PBR materials with realistic reflections on the glass and ceramics.
  • + Rich, detailed textures on the wood and sushi rice.
  • + Good centered composition with a clear isometric perspective.
  • Fails the 'mild garnish' requirement by adding many extra assets like a teapot, soy sauce bottle, and cups.
  • Stylistically leans more towards realism than the requested 'cartoon' scene.
  • The text 'SUSHI' is relatively small compared to 'JAPAN'.

Verdict: FLUX.2 [max] followed the stylistic instructions much more accurately, providing the requested 'cartoon' look with soft textures and minimal garnish. GPT Image 1.5 produced a much more cluttered and realistic scene that ignored the 'cartoon' and 'minimal' keywords, despite having superior material rendering.

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
FLUX.2 [max]
GPT Image 1.5
0% wins 25% ties 75% wins

AI Judge Analysis

FLUX.2 [max]

  • + Successfully incorporates all elements: TV anchor desk, dogs, and hockey background.
  • + The character maintains a distinct resemblance to the person in the source image in a cartoon style.
  • + Clean vector-style illustration with high clarity.
  • The 'caricature' style is more of a generic avatar/clipart style rather than an exaggerated caricature.
  • The dogs are repetitive in design.

GPT Image 1.5

  • + Excellent caricature style with exaggerated features that still strongly resemble the source subject.
  • + Highly creative integration of the dog wearing a hockey helmet and the 'Breaking News' ticker.
  • + Rich, vibrant colors and dynamic composition that feels energetic and humorous.
  • Minor text rendering issues on the microphone and ticker, though largely legible.
  • The transition between the desk and the character's body is slightly cluttered.

Verdict: GPT Image 1.5 is the clear winner as it delivered a true humorous caricature with exaggerated features while maintaining a striking resemblance to the source image. Although FLUX.2 [max] adhered to all prompt instructions, its style felt more like a flat digital illustration than the requested 'exaggerated and humorous' caricature.

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.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Excellent composition with a sense of movement as the animals run through the field.
  • + Very high level of realism in the fur texture and the dew drops on the grass.
  • + Correct anatomical rendering of all four requested animals with distinct features.
  • The lighting is a bit hazy, which slightly reduces the 'pop' of the subjects.
  • The butterflies look a bit static compared to the dynamic movement of the animals.

GPT Image 1.5

  • + Vibrant, warm lighting with very strong god rays and 'glittery' dew effects.
  • + Captures the 'tumbling' and 'joyful' vibe perfectly with expressive facial expressions.
  • + Extremely cute and stylized to emphasize the 'big expressive eyes' requested.
  • The puppy's left paw has five toes with prominent black pads that look slightly unnatural.
  • The animals are crowded together in a way that feels a bit less realistic than the spacing in Model A.
  • The fox's anatomy is slightly more generic/dog-like than the fox in Model A.

Verdict: FLUX.2 [max] provides a more sophisticated and realistic 8K masterpiece with better anatomical accuracy and a beautiful sense of depth. GPT Image 1.5 wins on pure emotional 'wholesomeness' and lighting effects, but it suffers from minor AI artifacts in the paws and a slightly more cluttered composition. FLUX.2 [max] is preferred for its cleaner technical execution and realistic integration of the four animals into the environment.

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
FLUX.2 [max]
GPT Image 1.5
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [max]

  • + Perfectly preserves the composition and poses of the original meme.
  • + Art style is very close to authentic hand-drawn anime with subtle paper textures.
  • + Maintains specific clothing details like the plaid pattern on the shirt accurately.

GPT Image 1.5

  • + Excellent soft, dreamy lighting and pastel color palette.
  • + Achieves a high level of aesthetic 'magic' associated with Ghibli backgrounds.
  • + Captures the emotional expressions of the characters very well in an illustrative style.
  • The man's hand is oddly merged with his side/hip.
  • The girl in the red dress has much longer, more voluminous hair than the source image.

Verdict: Both models did an excellent job translating the 'distracted boyfriend' meme into a Ghibli-inspired style. FLUX.2 [max] is the winner for its incredible preservation of the source image's geometry and details, whereas GPT Image 1.5 took more creative liberties with the hair and lighting that slightly drifted from the original composition.

Golden Hour Stroll

Image Editing
Edit instruction

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

Before After
FLUX.2 [max]
Before After
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [max]

  • + Successfully added hair blowing in the wind while maintaining hair texture consistency.
  • + Added floating leaves that appear somewhat natural to the scene.
  • + Preserved the identity of the woman and the dog almost perfectly.
  • The 'flying' leaves are static and lack motion blur, feeling more like they are pasted on top.
  • A weird artifact appears on the woman's left hand where her fingers now look distorted.

GPT Image 1.5

  • + Excellent addition of dynamic motion through motion-blurred leaves.
  • + The hair blowing looks very natural and follows a logical wind direction.
  • + The overall lighting and environment feel more 'energetic' as requested.
  • Slightly altered the woman's facial features compared to the source.
  • Several leaves overlap the dog and the woman's clothes in a way that looks like a digital overlay.

Verdict: Both models followed the instructions well, but GPT Image 1.5 captured the 'energetic and lively' feel much better by incorporating motion blur on the flying leaves. While FLUX.2 [max] preserved the source image details more accurately (aside from a hand glitch), its 'flying' leaves feel too static and disconnected from the 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.”

FLUX.2 [max]
GPT Image 1.5

AI Judge Analysis

FLUX.2 [max]

  • + Perfect text rendering for both name and date
  • + Accurately follows the light background requirement with subtle texture
  • + Excellent vector emblem composition that feels authentic for a vintage logo
  • The steam lines are a bit thin and could be more prominent

GPT Image 1.5

  • + Strong artistic texture on the cloche dome
  • + Dynamic typography style
  • Failed the light background prompt, using a pitch-black background instead
  • Rendering of the text 'FLORIAN' has minor inconsistencies in letter weight
  • The vector look is slightly more illustrative than a professional logo emblem

Verdict: FLUX.2 [max] followed the prompt much more accurately, especially regarding the light background and subtle texture requirements. While GPT Image 1.5 has an interesting artistic style, its failure to provide a light background and slight issues with font consistency make FLUX.2 [max] the clear winner for a professional logo design.

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.”

FLUX.2 [max]
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [max]

  • + Excellent typography with correct spellings for all names and labels.
  • + Well-organized grid layout following the requested sequential steps.
  • + Very clean flat-vector style with consistent iconography.
  • The sequence is slightly out of order, placing Lunar Orbit before Translunar.
  • Included yellow in the palette which wasn't specifically requested but fits the lunar module theme.
  • The 'Tranquiity' label has a minor typo.

GPT Image 1.5

  • + Strong NASA-inspired color palette usage and dramatic vector styling.
  • + Correct chronological ordering of all mission steps.
  • + High-quality illustrations of the Saturn V and Lunar Module.
  • The image is cropped at the top, cutting off the title.
  • The 'Translunar' step features a rocket rather than a trajectory arc icon as requested.
  • The 'Earth Orbit' step shows the Earth from the same perspective as the launch step, feeling redundant.

Verdict: FLUX.2 [max] produced a much more professional and complete infographic layout with clear labels and a polished aesthetic, despite an ordering error in the steps. GPT Image 1.5 followed the chronological order better but ultimately failed as a poster due to the top being cut off and less consistent icon styles.

Next steps

Explore each model