Head to head
Esc

Models · slot A

to navigate to pick

FLUX.1 [dev] Black Forest Labs GPT Image 1.5 OpenAI

Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.

FLUX.1 [dev]

24.6 arena score

#16 of 62 in Text-to-Image

Skill signature · Text-to-Image

GPT Image 1.5

27.1 arena score

#6 of 62 in Text-to-Image

Top 3 in Image Editing
Vote tally

Where the votes landed

FLUX.1 [dev]

0.0%

win rate

Ties

0.0%

GPT Image 1.5

100.0%

win rate

0.0% 0.0% ties 100.0%
Shared challenges 13

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent photorealism and texture on the leather book cover
  • + Physically plausible lighting and high-quality refractions
  • + Deep color depth and professional bokeh effect
  • The glass object is more of a thick-walled frame than a simple hollow cube
  • The sphere appears to be floating rather than resting on the surface

GPT Image 1.5

  • + Perfect adherence to the geometry of a glass cube
  • + The blue sphere is correctly resting on the bottom surface of the cube
  • + Clear visibility of the plant through the glass
  • Slightly lower image resolution and graininess in the background
  • The book edges look a bit more synthetic compared to Model A

Verdict: Both models followed the prompt instructions perfectly. FLUX.1 [dev] produced a more high-end, photographic image with superior textures, though it interpreted the cube as a more abstract glass structure. GPT Image 1.5 adhered more strictly to the literal physics of the objects (sphere resting on the base), making it a tie depending on whether the user prefers artistic quality or structural literalism.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent shallow depth of field and bokeh
  • + Accurate 50mm lens perspective
  • + Convincing cinematic light rain effects
  • The man is holding the handlebars rather than repairing the bike
  • The bike lacks a kickstand or support, making its standing position unrealistic

GPT Image 1.5

  • + Stronger narrative adherence with the man actually crouching to repair the mechanical parts
  • + Includes a tool tray and rag for added realism
  • + Excellent water droplets and surface texture on the bike and pavement
  • Motion blur on the passing car is more of a static blur rather than capturing a sense of speed
  • The composition is a bit tighter than specified for 'imperfect framing'

Verdict: While FLUX.1 [dev] captures a beautiful cinematic aesthetic with light rain, it fails the primary action of 'repairing' the bicycle, depicting the subject just standing next to it. GPT Image 1.5 provides a much more convincing and detailed scene of an actual repair with tools, better skin textures, and wet surface details, making it the more successful interpretation of the prompt.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Exceptional eye realism and facial lighting
  • + Very clean, high-resolution aesthetic
  • + Beautiful bokeh effect on the background sparks
  • Missed the 'battle-worn' and 'scarred' descriptors, appearing too clean
  • Hair beads are absent
  • Armor is relatively plain with minimal engraving

GPT Image 1.5

  • + Excellent adherence to the 'battle-worn' and 'scars' prompt elements
  • + Highly detailed engraving on the plate armor
  • + Successfully included hair beads and complex texture on leather and cloth
  • Image has a slightly over-sharpened, high-contrast digital look
  • Armor engravings look somewhat messy/incoherent on close inspection

Verdict: GPT Image 1.5 adhered much better to the prompt's specific details, capturing the battle-worn skin, scars, and braided beads that FLUX.1 [dev] omitted. While FLUX.1 [dev] produced a cleaner and more realistic character portrait, it failed to fulfill the 'battle-worn' and 'ornate' aspects of the request.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent minimalist graphic design with elegant curves
  • + Very clean professional layout
  • + Strong font hierarchy and use of negative space
  • Lacks the requested grid layout for photos
  • Most text is illegible placeholder gibberish
  • Missing a distinct Pizza category section

GPT Image 1.5

  • + Perfect adherence to the grid layout requirement
  • + Exceptional text legibility and realistic menu content
  • + Fully includes all three requested categories (Appetizers, Pizza, Mains)
  • Layout feels slightly more crowded compared to the 'minimalist' prompt
  • Colorful header accents are bit simplistic compared to Model A

Verdict: GPT Image 1.5 is the clear winner as it precisely follows all elements of the prompt, including the grid layout and specific category sections, while maintaining perfect text legibility. FLUX.1 [dev] produces a beautiful graphic design, but fails to implement a grid for the photos and contains mostly illegible text.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Clean layout with well-defined product separation
  • + Accurate photorealistic textures on the burger components
  • + Excellent formatting of the price and secondary text
  • Failed to include the primary title 'MAGIC BURGER'
  • Background lacks the 'fiery' intensity requested, appearing more like small flames
  • Missing the starburst for the price

GPT Image 1.5

  • + Successfully integrated all requested text with the correct fiery glowing effect
  • + Captured a high sense of motion and dynamic energy
  • + Features the required starburst for the price and a dramatic fiery background
  • The composition is slightly cluttered with many small debris particles
  • The burger bun on top looks a bit too wet/drippy compared to a standard ad

Verdict: GPT Image 1.5 followed the prompt much more accurately by including all requested text elements, the starburst, and the specific fiery aesthetic. While FLUX.1 [dev] produced a very clean and realistic food image, it completely missed the 'MAGIC BURGER' title and several specific design requirements.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent text legibility and accuracy of the specific menu items.
  • + Good variety in lettering styles between the title and the body.
  • The text looks more like a digital font overlay than real chalk strokes.
  • The 'handwriting' is too clean and uniform, lacking the requested natural variations and chalk texture.

GPT Image 1.5

  • + Superb chalk texture with realistic smudging and dusting on the board.
  • + Flawless adherence to the 'handwritten' request with natural variations in slant and character width.
  • + Correctly followed the cursive requirement for the title header.
  • The text is slightly harder to read due to the textured realism.
  • Wait for it—not necessarily a con, but the board is less 'centered' than Model A.

Verdict: While FLUX.1 [dev] produced very clear and accurate text, it failed the stylistic requirement of looking like authentic chalk, appearing more like a digital font. GPT Image 1.5 captured the chalk texture and handwriting style perfectly, including the requested cursive header and natural variations that make the board look truly hand-drawn.

The Reversed Rodeo

Text-to-Image

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

FLUX.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Successfully captures a surreal, dreamlike atmosphere with soft lighting
  • + Clean minimalist composition allows the subject to stand out
  • + Anatomically smooth rendering of the horse and suit
  • The horse's hind legs and hooves are anatomically deformed and messy
  • Failed the logic check: the prompt requested the horse riding the astronaut (horse on top)

GPT Image 1.5

  • + High level of detailed texture in the space suit and Lunar Lander
  • + Dynamic action with realistic dust/particle effects
  • + Cinematic lighting and rich background details
  • Failed the logic check: interpreted the prompt as the standard astronaut riding a horse
  • The scale of Saturn and Earth in the background is visually cluttered

Verdict: Both FLUX.1 [dev] and GPT Image 1.5 failed the specific logic constraint of the horse riding the astronaut rather than the other way around. FLUX.1 [dev] provides a more surreal and clean image, but it suffers from significant anatomical glitiches on the horse's legs, whereas GPT Image 1.5 is visually busier but much higher in technical detail and texture clarity.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent image clarity and high-resolution textures.
  • + Good lighting and cinematic atmosphere.
  • Composition error: the passenger is sitting in the front seat instead of the back seat.
  • The capybara's paws are not placed logically on the steering wheel.

GPT Image 1.5

  • + Correct composition with the passenger clearly in the back seat as requested.
  • + Authentic taxi driver accessory with the checkered trim on the cap.
  • + Natural placement of the capybara's paws on the steering wheel.
  • The passenger's face is slightly blurry and lacks detail compared to Model A.
  • The image has a more grainy, film-like texture which may slightly reduce clarity.

Verdict: Both models followed the complex prompt well, but Model B is the clear winner for spatial accuracy. While FLUX.1 [dev] produced a sharper image, it failed the simple spatial instruction by placing the passenger in the front seat, whereas GPT Image 1.5 correctly depicted the scene with the businesswoman in the back seat and a more convincing driver's pose for the capybara.

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.1 [dev]
GPT Image 1.5
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.1 [dev]

  • + Strong composition with a clean graphic design feel
  • + Vibrant and cinematic lighting on the jack-o-lantern
  • + Text is generally legible and correctly placed
  • Failed the main title text spelling with 'Falloween Rantcj' instead of 'Halloween Party Invitation'
  • The aesthetic is more like modern vector art than the requested 'vintage gothic parchment'
  • Contains minor text artifacts like 'Time: 7pm, 7pm'

GPT Image 1.5

  • + Perfect adherence to the 'vintage gothic parchment' texture and style
  • + Accurate spelling for all requested text, including the main title and banner
  • + Excellent inclusion of all requested details like spider webs, thorns, and twisted trees
  • The darker color palette makes some fine details in the background slightly muddy
  • The text on the scroll banner is a bit thin against the textured background

Verdict: GPT Image 1.5 is the clear winner as it perfectly captured the vintage gothic aesthetic and correctly rendered all the requested text, whereas FLUX.1 [dev] failed significantly on the main title's spelling. GPT Image 1.5 also followed the specific stylistic cues like 'parchment' and 'webs' much more effectively than the clean, vector-style output of FLUX.1 [dev].

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent soft, clay-like 3D miniature textures
  • + Clean and minimal aesthetic
  • + Graceful inclusion of decorative elements
  • Text rendering is poor with 'SUSH CATON' and messy kanji
  • Lower resolution/slight blur compared to Model B

GPT Image 1.5

  • + Perfect text rendering for both 'JAPAN' and 'SUSHI'
  • + Highly detailed 3D assets with rich PBR materials
  • + Stronger adherence to the '45° top-down isometric' perspective
  • Scene is slightly cluttered compared to the request for 'minimal garnish'

Verdict: GPT Image 1.5 followed the complex instructions, particularly the text requirements, much better than FLUX.1 [dev], which struggled with spelling and alignment. While FLUX.1 [dev] captured a softer, more artistic 3D style, GPT Image 1.5 provided a clearer, more professional diorama with better material definition.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent adherence to the request for bright, joyful lighting.
  • + Clean composition with consistent character styling.
  • Faces are highly stylized and cartoonish, failing the hyper-photorealistic requirement.
  • Failed to include a tabby kitten, instead showing two dogs and two fox/rabbit hybrids.
  • Anatomy is simplified and lacks realistic fur texture.

GPT Image 1.5

  • + Successfully achieves a hyper-photorealistic style with complex fur textures.
  • + Accurately includes all four requested animals: golden retriever, tabby kitten, bunny, and fox kit.
  • + Effective use of god rays and dew sparkles to create a magical atmosphere.
  • One of the fox kit's paws appears somewhat murky and poorly defined.
  • The composition is a bit crowded compared to the cleaner layout of Model A.

Verdict: GPT Image 1.5 is the clear winner as it followed the complex prompt requirements for specific animal types and achieved a hyper-photorealistic look. FLUX.1 [dev] produced a charming but overly cartoonish image that failed to include the tabby kitten and leaned toward a 3D-animation aesthetic rather than realism.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Features a light background as requested
  • + Clean vector emblem style with good balance
  • + Includes a banner for the text
  • Severely misspelled main brand name as 'Flarilaan'
  • Misspelled 'Restaurant' as 'Reseaurant'
  • Cloche is very small and lacks detail

GPT Image 1.5

  • + Perfect text rendering for both 'Caffè Florian' and 'Est. 1720'
  • + Beautifully detailed cloche dome with subtle texture
  • + Stronger retro aesthetic and professional typography
  • Ignored the 'light background' instruction, opting for black
  • Cloche lacks the requested 'steam' effect, showing only abstract shapes above it

Verdict: GPT Image 1.5 is the clear winner because it correctly spells the brand name and date, whereas FLUX.1 [dev] contains multiple significant typos. Although GPT Image 1.5 failed the background color requirement, its superior typography and illustration quality make it a much more usable logo.

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.1 [dev]
GPT Image 1.5

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent aesthetic adherence to the 'modern vector' and 'NASA-inspired' color palette.
  • + Sophisticated composition with a clean, centralized layout.
  • + High level of detail in the small iconography at the bottom.
  • Text rendering is mostly gibberish despite the clear typography style.
  • Iconography path is logically confusing and includes random planets like Saturn.
  • Fails to clearly delineate the 6 requested steps in sequence.

GPT Image 1.5

  • + Perfect adherence to the 6 requested steps with accurate labeling.
  • + Excellent text rendering of step titles and crew names.
  • + Clean, professional flat-vector blocks that are easy to read as an infographic.
  • Composition is a bit more rigid and standard than Model A.
  • Visual elements are slightly more 'cartoonish' compared to the requested sleek modern style.

Verdict: GPT Image 1.5 is the clear winner for its superior prompt adherence, correctly depicting all six requested mission steps with perfect text labeling. While FLUX.1 [dev] produced a more artistically sophisticated layout and better color harmony, its failure to render legible text or follow the specific logical sequence of the Apollo mission makes it less useful as an infographic.

Next steps

Explore each model