Black Forest Labs' aesthetically-tuned 12-billion parameter flow transformer optimized for high-quality images with incredible aesthetics, suitable for personal and commercial use
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 Krea [dev]
18.8
arena score
#48 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.1 Krea [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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent photorealistic lighting and shallow depth of field.
- + Highly realistic textures on the glass, book, and wooden surface.
- + Accurate reflection of the blue sphere within the glass.
- − The red book is slightly off-center and floating slightly above the glass edge.
- − The blue sphere appears more like a glass marble than a simple sphere.
GPT Image 1.5
- + Strong adherence to all spatial requirements in the prompt.
- + Clear visibility of the green plant through the glass panels.
- + Solid composition with the book perfectly aligned on top of the cube.
- − The image has a slightly smoothed, digital look compared to Model A.
- − The lighting feels a bit flatter and less natural than the competitive image.
Verdict: FLUX.1 Krea [dev] produces a significantly more realistic and aesthetically pleasing photograph with superior lighting and materials, though it has minor alignment issues with the book. GPT Image 1.5 follows the positioning instructions perfectly but lacks the photorealistic depth and texture quality found in the FLUX output. Overall, FLUX.1 Krea [dev] is the winner due to its professional photographic quality.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent depiction of motion blur on passing cars
- + Cinematic color grading and lighting
- + Captures the 'imperfect framing' request effectively
- − The subject is just holding the bike rather than repairing it
- − The bicycle anatomy is slightly warped near the front tire and handlebars
GPT Image 1.5
- + High adherence to the action of 'repairing' with a tool tray visible
- + Superior skin texture and fine detail rendering on the face and clothing
- + Realistic wet surface reflections and textures
- − Lacks the motion blur from passing cars requested in the prompt
- − Composition feels a bit more staged than 'candid street photo'
Verdict: GPT Image 1.5 provides a much more detailed and accurate interpretation of the 'repairing' action with superior texture work, whereas FLUX.1 Krea [dev] better captures the specific environmental requests like motion blur and imperfect street framing. GPT Image 1.5 is the preferred image due to its overall clarity and closer adherence to the central subject's activity.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent engraving details on the armor and plate texture
- + Strong adherence to the braided hair with beads requirement
- + Clean, cinematic composition with balanced bokeh
- − The 'scars' look more like fresh blood splatters than healed battle scars
- − Lighting is a bit flat across the face compared to the armor
GPT Image 1.5
- + Deeply realistic skin texture with lifelike eyes and subtle scars
- + Superior rendition of warm torchlight reflecting off metallic surfaces
- + Exceptional detail on the worn leather straps and cloth underlayer
- − The braids are a bit messy and less distinct than the ones in the other model
- − Composition is a very tight crops which loses some of the silhouette
Verdict: GPT Image 1.5 produces a more lifelike and gritty portrait that perfectly captures the lighting and texture requirements of the prompt, particularly the 'battle-worn' aesthetic and warm torchlight. FLUX.1 Krea [dev] has cleaner armor engravings and clearer braids, but its skin details feel more like makeup/paint than authentic scars and dirt.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Elegant mockup presentation with realistic shadows.
- + Large, diverse grid of colorful food photography.
- + Consistently clean, minimalist layout.
- − Text is largely illegible gibberish.
- − Repeated 'Appetizers' section instead of the requested 'Pizza' section.
GPT Image 1.5
- + Excellent text rendering with clear, legible English.
- + Strict adherence to section headers (Appetizers, Pizza, Mains).
- + High-quality, appetizing food photos that correspond to the menu items.
- − Layout is slightly more basic compared to a professional design software export.
- − Lack of a mockup presentation (e.g., paper texture or shadows) makes it look like a digital file.
Verdict: While FLUX.1 Krea [dev] produces a more aesthetically pleasing mockup, it fails significantly on text legibility and following the specific section prompts. GPT Image 1.5 provides a fully functional, highly legible menu design that perfectly matches the requested sections and typography 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.1 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Clean, professional advertisement layout
- + Excellent text legibility and font consistency
- + Photorealistic food styling
- − Failed to render the price inside a starburst
- − The 'exploded' effect is very conservative with components mostly stacked
- − Graphic design elements at the bottom contain gibberish text
GPT Image 1.5
- + Perfect adherence to the 'exploded burger' concept with high dynamic motion
- + Accurately rendered the price within a fiery starburst as requested
- + Consistent fiery, glowing effect applied to all text elements
- − The composition is a bit crowded/busy
- − Misses the 'limited time only' tagline in the main title area, relegated to the bottom
Verdict: GPT Image 1.5 follows the complex prompt much more accurately, successfully integrating the exploded burger physics, fiery aesthetic, and the specific starburst requirement for the price. While FLUX.1 Krea [dev] produces a very clean and realistic food shot, it fails on several specific prompt instructions like the starburst and the intensity of the 'exploded' effect.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent framing and realistic board texture
- + Clean, legible handwriting with professional café aesthetics
- − Included severe hallucinations and typos like 'Gruffle Mushroom' and 'with tcoman'
- − Failed to complete the 'Brown But...' prompt instruction correctly by adding extra garbled lines
GPT Image 1.5
- + Perfect text accuracy for all requested items
- + Superior chalk texture realism with smudges and dusty background
- + Followed the handwriting style instructions consistently across the board
- − The tight cropping loses the 'cozy café' context visible in the other model
Verdict: GPT Image 1.5 followed the prompt much more effectively, rendering the text exactly as requested without the significant spelling errors found in FLUX.1 Krea (dev). While FLUX.1 Krea (dev) has better framing and lighting, its inability to maintain text coherence—mistaking 'Truffle' for 'Gruffle' and adding nonsensical words—makes it less useful for this specific task.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.1 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Successfully followed the difficult spatial constraint of 'horse on top'
- + Clean, cinematic lighting and high resolution
- + Coherent blending of astronaut equipment and horse tack
- − The horse's anatomy becomes surreal/abstract where it merges with the white mechanical parts
- − Slightly less background detail compared to the competitor
GPT Image 1.5
- + Very high detail in the environment and textures
- + Dynamic composition with a sense of motion
- + Good rendering of the astronaut suit
- − Failed the primary prompt constraint of having the horse on top of the astronaut
Verdict: This was a test of prompt adherence regarding spatial relationships. FLUX.1 Krea [dev] followed the instruction to place the horse on top of the astronaut, whereas GPT Image 1.5 ignored the specific 'not vice versa' instruction and generated a standard astronaut riding a horse.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Includes the taxi rooftop sign visible through the window.
- + Clear, high-resolution rendering of the capybara's fur.
- − Includes a third person/extra head in the front passenger seat which was not requested.
- − The capybara's paws look more like mammalian hands rather than webbed capybara feet.
- − The perspective feels slightly distant compared to the requested interior scene.
GPT Image 1.5
- + Perfect framing that creates a sense of being inside the vehicle with the driver.
- + Accurate capybara anatomy for the paws on the steering wheel.
- + Excellent lighting and photorealistic background bokeh.
- − Slightly less clarity on the text of the businesswoman's phone.
Verdict: GPT Image 1.5 is the clear winner as it correctly followed the instructions without adding extra characters, whereas FLUX.1 Krea [dev] added a third person in the front seat. GPT Image 1.5 also provided a more immersive 'inside the car' perspective and more realistic capybara paws.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent graphic design layout with clean borders
- + Modern, polished illustration style
- + High contrast lighting on the jack-o-lantern
- − Serious text errors including 'Pasty Halloween' and repeated words
- − Failed to follow the date and location text instructions accurately
- − Parchment texture is less evident than in the competitor
GPT Image 1.5
- + Perfect text rendering for all requested details including date, time, and location
- + Rich, atmospheric vintage parchment texture with thorns and webs
- + Superior adherence to the gothic aesthetic and prompt details
- − The jack-o-lantern's glow is a bit muddy compared to the rest of the scene
- − Composition is slightly crowded toward the bottom
Verdict: GPT Image 1.5 is the clear winner here as it successfully rendered all the requested text with zero errors, whereas FLUX.1 Krea [dev] failed significantly on the typography. GPT Image 1.5 also captured the specific 'vintage gothic' feel much better through its use of texture and the thorny border.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent adherence to the 'minimal garnish' and 'clean' aesthetic requested.
- + High quality 3D rendering with soft, appealing lighting and textures.
- + Perfectly centered and balanced composition.
- − The 'flag icon' is rendered as a physical object on the sushi rather than part of the top-center text layout.
- − Typography is a bit bubbly and less 'bold' compared to model B.
GPT Image 1.5
- + Highly detailed textures on the wood and materials.
- + Followed the specific three-line text/icon arrangement at the top accurately.
- + Very clear 45° isometric perspective.
- − Failed to provide 'minimal' garnish and plate, creating a very complex scene instead.
- − The diorama base is much larger and more cluttered than requested.
Verdict: FLUX.1 Krea (dev) followed the aesthetic constraints of 'minimal' and 'clean' much better than GPT Image 1.5, which ignored those keywords to create a cluttered, albeit detailed, scene. While GPT Image 1.5 placed the flag icon more accurately according to the prompt's layout description, FLUX.1 Krea's output is closer to the intended 'miniature diorama' style with superior visual balance.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent adherence to lighting requests with beautiful god rays and soft bokeh.
- + Clean character separation and sharp, high-quality textures on the fur.
- + Successfully captures the 'chasing' and 'tumbling' motion described.
- − Prompt adherence failure: missed the baby bunny entirely.
- − Included two tabby kittens instead of just one.
GPT Image 1.5
- + Perfect prompt adherence, including all four specific animals: puppy, kitten, bunny, and fox kit.
- + Dynamic composition with the 'tumbling' together aspect clearly depicted.
- + Beautifully rendered dew sparkles and atmospheric golden light.
- − Slight anatomical awkwardness on the kitten's front paws.
- − The fox's eyes appear slightly less realistic/very doll-like compared to the other animals.
Verdict: GPT Image 1.5 is the winner because it successfully included all four requested animals, whereas FLUX.1 Krea [dev] missed the bunny and duplicated the kitten. While FLUX.1 Krea [dev] had slightly cleaner fur textures, GPT Image 1.5 captured the 'tumbling' interaction and 'dew sparkles' much better, adhering more closely to the complex prompt.
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 Krea [dev]
GPT Image 1.5
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent hand-drawn etching style that fits the vintage aesthetic.
- + Beautiful use of parchment-style texture and negative space.
- + Accurate date and cloche integration.
- − The main text 'Florian' is misspelled as 'FLANDR'IN'.
- − The typography composition is a bit cluttered with the banner overlapping the text.
GPT Image 1.5
- + Perfect spelling of the requested name 'Caffè Florian'.
- + Clean layout with the banner positioned correctly at the base.
- + Strong vector feel with good contrast and grain.
- − Ignored the request for a 'light background', providing a black background instead.
- − The cloche dome rendering is less artistic compared to the etching style of Model A.
Verdict: GPT Image 1.5 is the winner primarily for its ability to follow the text requirements, correctly spelling 'Caffè Florian' while FLUX.1 Krea [dev] failed the spelling. Although FLUX.1 Krea [dev] captured the vintage texture and artistic style much better, the spelling error and the failure to provide a clean logo layout make it less practical than the GPT Image 1.5 output.
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 Krea [dev]
GPT Image 1.5
AI Judge Analysis
FLUX.1 Krea [dev]
- + Successfully captures a modern 'flat' vector aesthetic for the base background.
- + Legible main header text.
- − Fails to follow the iconography sequence entirely, showing random figures and planets.
- − The logic and flow of the infographic are nonsensical and messy.
- − Garbled text and incorrect number count.
GPT Image 1.5
- + Follows the specific 6-step infographic sequence perfectly with accurate iconography.
- + Excellent typography with high legibility and correct spelling of astronaut names and 'Tranquillity'.
- + Clean, professional layout that matches the desired NASA color palette and vector style.
- − The 'Descent' and 'Landing' boxes have slightly repetitive visual compositions.
- − Minor green tints in the Earth icon deviate slightly from the 'muted red/navy' constraint.
Verdict: GPT Image 1.5 is the clear winner as it followed all instructions, including the specific 6-step sequence and requested iconography. FLUX.1 Krea [dev] failed to create a logical infographic, producing jumbled icons and garbled text that did not relate to the Apollo 11 mission steps.
Next steps
Explore each model
OpenAI's state-of-the-art image generation model with better instruction following and adherence to prompts