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]
#48 of 62 in Text-to-Image
Wan 2.5 (Preview)
#28 of 62 in Text-to-Image
Where the votes landed
FLUX.1 Krea [dev]
0%
win rate
Ties
0%
Wan 2.5 (Preview)
0%
win rate
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent photo-realistic lighting and reflections on the sphere
- + The glass cube has realistic thickness and distorted reflections
- + Subtle and sophisticated composition
- − The plant is very dark and its presence 'through the glass' is less defined than the other model
Wan 2.5 (Preview)
- + Perfect adherence to the requirement of the plant being visible through the glass
- + Bright, vibrant colors and sharp details
- + Distinct and clear rendering of the 'small blue sphere'
- − The lighting is a bit harsh with some floating dust artifacts that may look like low-quality noise
- − The reflection of the sphere on the bottom of the cube doesn't quite match its matte texture
Verdict: Both models followed the complex spatial instructions perfectly. FLUX.1 Krea [dev] produced a more believable, high-end photographic look, but Wan 2.5 (Preview) more clearly demonstrated the 'plant' being viewed through the glass cube as requested in the prompt. FLUX.1 Krea [dev] is the winner for its superior handling of material properties and lighting.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Successfully captures motion blur from passing cars
- + Realistic atmospheric lighting and reflections
- + Captures an 'imperfect' street photography framing well
- − The man is walking/holding the bike rather than 'repairing' it
- − Lower overall sharpness and detail on the subject face
Wan 2.5 (Preview)
- + Closer adherence to the 'repairing' action with tools visible
- + Excellent skin texture and fine detail on the face and hands
- + Very high visual clarity and beautiful reflections
- − Lacks the requested 'motion blur from passing cars'
- − The red paint on the bike looks slightly too clean and saturated for the 'no stylization' request
Verdict: Wan 2.5 (Preview) produced a higher quality image with superior skin details and better adherence to the specific action of repairing the bike. While FLUX.1 Krea [dev] captured the atmospheric elements like motion blur and candid framing better, it failed to show the man actually repairing the bicycle.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent depiction of ornate engraved plate armor with high-contrast texturing.
- + Sophisticated use of lighting with subtle warm reflections following the facial structure.
- + Dramatic atmosphere with effective use of bokeh sparks.
- − The 'braided hair with small beads' is very subtle and mostly limited to the hairline.
- − The armor engraving, while beautiful, feels slightly repetitive/pattern-like.
Wan 2.5 (Preview)
- + Outstanding facial detailing including bloodshot eyes, pores, and realistic dirt/scars.
- + Stronger adherence to the braided hair and beads aspect of the prompt.
- + Superior texture work on the leather straps and frayed cloth underlayers.
- − The torch flame in the background looks slightly artificial compared to the subject.
- − The character looks a bit young for the 'battle-worn' description compared to Model A.
Verdict: Both models followed the complex prompt with high fidelity, but Wan 2.5 (Preview) produced a more varied and tactile image with superior detailing on the leather and cloth elements. FLUX.1 Krea [dev] created a more classically handsome, cinematic portrait with impressive metalwork, but Wan 2.5's technical execution of skin textures and multiple layers of clothing makes it the more realistic output.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent structure that replicates a real-world flyer or menu page.
- + Clear, bold typography with a professional hierarchy of information.
- + Realistic food photography that feels cohesive and appetizing.
- − Text contains several spelling errors (e.g., 'Desiorn', 'Maines').
- − Includes 'Appetizers' twice as a header instead of distinguishing the second section.
Wan 2.5 (Preview)
- + Accurately portrays a grid-based layout for all three requested sections.
- + High-quality, isolated food photography with consistent lighting.
- + Good use of vibrant color accents to separate categories.
- − Layout feels more like a conceptual web page than a practical printed menu.
- − Heavy spelling errors in titles (e.g., 'Restormalit Menue', 'Foomohnrieus').
- − Text rendering for body descriptions is very wobbly and illegible.
Verdict: FLUX.1 Krea [dev] produces a much more convincing and professional menu layout that feels ready for use in a casual dining setting, despite some minor spelling errors. While Wan 2.5 (Preview) follows the grid layout request strictly, its text rendering and overall composition feel more like a rough draft or a website wireframe than a functional physical menu.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Clean, legible typography for the main title.
- + Strong use of photographic depth of field in the burger textures.
- + Effective integration of the fiery background elements.
- − The burger components are stacked rather than 'exploded' or suspended as requested.
- − Includes distracting gibberish text at the bottom.
- − The price element lacks the requested glowing fiery effect.
Wan 2.5 (Preview)
- + Excellent adherence to the 'exploded' dynamic prompt with widely separated components.
- + Highly creative typography with a melting, fiery aesthetic.
- + Perfect execution of the price starburst with glowing effects.
- − The price text is slightly less crisp than the main title.
- − Includes several repetitive tomato slices that feel slightly cluttered.
Verdict: Wan 2.5 (Preview) is the winner for its superior interpretation of 'exploded' components and creative, fiery typography that perfectly matches the ad's theme. While FLUX.1 Krea [dev] has sharp photography, it produces a standard stacked burger and fails to correctly implement the specific 'starburst' and 'glow' requests for the price.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Features a very realistic chalk smudge texture in the corners of the board.
- + Attempts more complex cursive elements as requested in the prompt.
- + Follows the request for a straight-on composition.
- − Includes several spelling errors and hallucinates extra menu items ('Gruffle Mushroom', 'Chvrcuations').
- − The handwriting looks more like a digital monoline font than organic chalk strokes.
- − Text rendering becomes somewhat garbled toward the bottom.
Wan 2.5 (Preview)
- + Excellent text accuracy with almost zero spelling errors on the requested items.
- + The handwriting has varying thickness and pressure, making it look much more like real chalk.
- + Naturally completes the 'Brown But...' prompt fragment into 'Brown Butter Chocolate Chip Cookies'.
- − The perspective is slightly angled rather than being a centered shot of the menu board.
- − Relies more on a print-script than an 'elegant cursive' for the title.
- − The layout is slightly cramped at the bottom.
Verdict: Wan 2.5 (Preview) significantly outperforms FLUX.1 Krea [dev] in both text rendering accuracy and the physical realism of the chalk texture. While Image A struggles with spelling and creates nonsensical words, Image B correctly interprets the prompt items and maintains a consistent, high-quality handwriting style.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent adherence to the 'surreal' aspect of the prompt with the cyborg-like horse construction.
- + High cinematic quality with dramatic lighting and a clear focal point.
- + Successfully interprets the unusual request of horse and astronaut positioning in a creative way.
- − The anatomy of the horse becomes very muddled where it merges with the mechanical parts.
- − The lighting on the horse's head is a bit harsh compared to the rest of the scene.
Wan 2.5 (Preview)
- + Vibrant colors and a very high level of detail in the astronaut suit and horse's coat.
- + Strong composition with dynamic movement and a beautiful celestial background.
- + Clean rendering without obvious architectural glitches.
- − Failed the specific prompt instruction 'horse on top, not vice versa'.
- − The inclusion of dust/debris at the bottom suggests a ground surface, which slightly contradicts the 'in space' setting.
Verdict: FLUX.1 Krea [dev] is the clear winner because it followed the difficult negative/spatial constraint of having the 'horse on top' (interpreting it as the horse being the dominant, larger-than-life structure carrying the astronaut or integrated with space tech). Wan 2.5 (Preview) produced a higher quality, more vibrant image, but completely ignored the core instruction to flip the usual rider relationship, resulting in a standard 'astronaut riding a horse' image.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent texture on the capybara's fur
- + Good interpretation of 'bored expression' on the passenger
- − An extra person is confusingly appearing in the front passenger seat
- − The capybara's hands look more like paws but the grip on the wheel is less defined than model B
- − The 'TAXI' text on the hat is a bit basic
Wan 2.5 (Preview)
- + Perfect adherence to the single passenger in the back seat requirement
- + High level of detail on the taxi exterior and rain drops
- + More realistic driver's cap style
- − The capybara's paws look slightly creepy/humanoid with long claws
- − The text on the taxi roof sign is distorted/gibberish
Verdict: Wan 2.5 (Preview) followed the prompt instructions much better by including only one passenger in the back seat and capturing the rainy Manhattan atmosphere. FLUX.1 Krea [dev] had better fur texture but failed by adding an unnecessary extra person in the front seat, which cluttered the composition.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent central jack-o-lantern illustration
- + Sophisticated gothic border design with webs
- + Great atmospheric lighting with silhouettes
- − Significant text errors including 'Pasty Halloween' and repeated words in the banner
- − Layout is a bit top-heavy with the large header text
Wan 2.5 (Preview)
- + Perfect text rendering for all requested fields including the specific date and location
- + Highly detailed thorns and spider web border as requested
- + Beautiful cinematic lighting and high-quality textures on the pumpkin
- − The 'scroll banner' is slightly merged into the parchment background
- − The arrangement of the thorns is quite dense, bordering on cluttered
Verdict: Wan 2.5 (Preview) significantly outperforms FLUX.1 Krea in this task by following all text instructions perfectly, including the specific date and location which FLUX.1 garbled. While FLUX.1 Krea has a strong artistic atmosphere, the typos like 'Pasty Halloween' and 'You are are' make the invitation unusable compared to the polished and accurate output from Wan 2.5.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent adherence to the 45-degree isometric perspective and diorama base request.
- + Perfect text rendering with high clarity.
- + The flag icon is physically integrated into the scene, enhancing the miniature feel.
- − The lighting is a bit harsh with deep, sharp shadows compared to the 'gentle lighting' request.
- − Texture of the salmon is slightly less appetizing and looks more like plastic than a refined 3D render.
Wan 2.5 (Preview)
- + Beautiful soft lighting and refined, rounded 3D textures that match the 'cartoon' and 'PBR' style perfectly.
- + Excellent composition with a professional graphic design feel.
- + High-quality text rendering and inclusion of all prompt elements.
- − The perspective is more of a front-view/low-angle rather than the requested 45-degree top-down isometric view.
- − Does not include the specific 'raised diorama base', instead using a circular plate stand.
Verdict: FLUX.1 Krea [dev] followed the technical spatial instructions much better, delivering a true isometric top-down scene on a square diorama base. However, Wan 2.5 (Preview) produced a more visually appealing image with superior lighting and texture quality, even though it ignored the specific perspective and base shape requirements.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent soft lighting and atmospheric god rays
- + High-quality fur textures and realistic anatomy
- + Subtle, natural-looking interaction with the environment
- − Failed to include the baby bunny
- − Included two kittens instead of one
- − The fox kit has slightly unusual dark markings on its chest
Wan 2.5 (Preview)
- + Successfully included all four requested animals (dog, cat, bunny, fox)
- + Captured the 'hyper-photorealistic' and 'dew sparkles' descriptors very literally
- + Dynamic posing gives a strong sense of chasing and tumbling
- − The fox's eyes appear unnaturally blue and glowing
- − Floating water droplets look a bit like CGI artifacts rather than realistic dew
- − The kitten's tail is oddly positioned
Verdict: Wan 2.5 (Preview) followed the complex prompt more accurately by including all four specific animals, whereas FLUX.1 Krea [dev] missed the bunny and duplicated the kitten. However, FLUX.1 Krea [dev] achieved a much more professional, soft photographic look with superior lighting, while Wan 2.5 (Preview) felt slightly more like a rendered illustration due to the over-saturated eyes and floating droplets.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent hand-drawn etching style that conveys a genuine vintage feel.
- + Sophisticated use of texture on the background.
- + Correct date rendering on the banner.
- − Spelling error in the main name, rendering it as 'Café Flanelín' instead of 'Caffè Florian'.
- − The steam element is somewhat singular and heavy-handed compared to the rest of the illustration.
Wan 2.5 (Preview)
- + Perfect text rendering for both 'Caffè Florian' and 'Est. 1720'.
- + Clean vector-style emblem that follows the minimalist part of the prompt well.
- + Balanced composition with an elegant border.
- − The paper texture looks a bit like a stock photo background rather than a subtle integrated texture.
- − The shading on the cloche is a bit generic.
Verdict: While FLUX.1 Krea (dev) captures a much more authentic vintage aesthetic with its detailed etching style, it fails the primary text requirement by misspelling the restaurant name. Wan 2.5 (Preview) provides a clean, professional vector logo that follows every part of the prompt perfectly, including the exact spelling and date.
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.”
AI Judge Analysis
FLUX.1 Krea [dev]
- + Strong 'Apollo 11' heading font
- + Adheres well to the simplified minimalist request for flat vector icons
- − Nonsensical iconography for steps (e.g., Saturn-like planet, 4 people instead of 3)
- − Fails to follow textual instructions for the 6 specific steps
- − Poor text rendering for labels
Wan 2.5 (Preview)
- + Excellent adherence to specific mission step instructions
- + Perfectly rendered, legible text for mission labels and names
- + Accurate technical illustrations of the Saturn V and Lunar Module
- − Slightly more complex than the 'flat vector' request with some detailed shading
- − The trajectory lines are slightly cluttered
Verdict: Wan 2.5 (Preview) is the clear winner as it accurately followed all six instructional steps with high-quality, legible text and relevant iconography (Saturn V, Lunar Module, and correct crew count). FLUX.1 Krea [dev] failed to generate relevant icons for the steps, showing generic space shapes and garbled text that did not relate to the Apollo 11 mission logic.
Explore each model
Alibaba's text-to-image and image-to-image generation model from the Wan AI suite, offering high-quality visual generation capabilities