Black Forest Labs' 12-billion parameter multimodal flow transformer for in-context image generation and editing with character consistency, typography handling, and commercial-ready quality
Settled by community votes across 17 shared challenges, with an AI judge weighing in on each.
FLUX.1 Kontext [pro]
#41 of 62 in Text-to-Image
Wan 2.5 (Preview)
#27 of 62 in Text-to-Image
Where the votes landed
FLUX.1 Kontext [pro]
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 Kontext [pro]
- + Excellent adherence to the glass tray/cube structure and lighting direction.
- + Very clean, minimalist aesthetic with realistic texture on the table surface.
- + Perfect placement of the plant in the background as requested.
- − The glass 'cube' feels more like a light frame or tray than a solid glass object.
- − The blue sphere has a slightly fuzzy or felt-like texture which may not be expected.
Wan 2.5 (Preview)
- + Beautiful glass refractive effects and reflections on the sphere.
- + Higher level of detail and character in the textures of the book and plant.
- + Dynamic lighting with visible dust motes adds to the atmosphere.
- − The book's placement is physically awkward, appearing to balance on a thin edge rather than a flat surface.
- − The glass cube ends abruptly at the bottom, looking more like an open box than a closed cube.
Verdict: Both models followed the prompt instructions perfectly, including the specific spatial relationships between objects. FLUX.1 Kontext [pro] produced a cleaner, more structurally sound image with a realistic wood grain, while Wan 2.5 (Preview) offered more impressive lighting effects and material textures at the cost of some physical coherence in the glass and book placement.
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 Kontext [pro]
- + Excellent skin texture and facial realism.
- + Atmospheric rain effects and bokeh are very natural.
- + Captures the 'imperfect framing' and cinematic feel requested.
- − The man appears to be riding or holding the bike rather than 'repairing' it.
- − The bicycle's anatomy is slightly warped at the handlebars.
Wan 2.5 (Preview)
- + Strong adherence to the 'repairing' action with tools on the ground.
- + Great reflections on the wet pavement.
- + Accurate red bicycle color and realistic crouching pose.
- − The rain looks slightly like static or vertical lines rather than falling droplets.
- − Skin texture is a bit smooth and lacks the requested 'natural' imperfections compared to the other model.
Verdict: FLUX.1 Kontext [pro] creates a much more convincing 'candid' and cinematic atmosphere with superior skin textures, but it fails to show the man actually repairing the bike. Wan 2.5 (Preview) follows the activity prompt much better, showing tools and specialized mechanics, though the image quality feels slightly more synthetic.
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 Kontext [pro]
- + Excellent skin texture with realistic fine pores and faint scarring
- + Sophisticated warm lighting that feels integrated into the scene
- + High-quality rendering of the engraved metal and fabric textures
- − Hair beads are very subtle and almost unnoticeable
- − The 'battle-worn' aspect is quite clean compared to the prompt
Wan 2.5 (Preview)
- + Strong adherence to the 'beads in hair' and 'dirt/scars' requirements
- + Beautifully detailed engraving on the pauldrons and visible chainmail
- + Great use of bokeh and actual torchlight within the frame
- − The character looks slightly younger than a typical 'battle-worn paladin'
- − Skin rendering is slightly smoother/waxy compared to Model A
Verdict: Both models performed exceptionally well on this complex prompt. FLUX.1 Kontext [pro] offers a more mature, photorealistic portrait with superior skin texture, while Wan 2.5 (Preview) followed specific instructions like the hair beads and heavier dirt/grime much more closely. Wan 2.5 is the winner for its better balance of all prompt elements and intricate armor detailing.
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 Kontext [pro]
- + Excellent typography rendering with readable header text.
- + Logical layout where food items correspond to their respective categories.
- + Clean, professional white space usage typical of a real menu.
- − The 'Pizza' and 'Mains' sections both show images of pizza, which is redundant.
- − The food photos are inconsistently shaped and lack a formal grid feel.
Wan 2.5 (Preview)
- + Perfectly follows the grid layout request for food photos.
- + More vibrant and consistent food photography across all items.
- + Better adherence to the 'colorful' and 'vibrant accents' part of the prompt.
- − Text rendering is poor and contains gibberish/artifacts.
- − The logical hierarchy is broken, as images appear above category headers that don't match (e.g., Pizza under Appetizers).
Verdict: FLUX.1 Kontext [pro] produces a more functional design with legible text and a realistic layout, though it struggles with variety in the food imagery. Wan 2.5 (Preview) excels at the requested 'grid' aesthetic and visual vibrance, but the text is unreadable and the categorization of the food items is nonsensical. FLUX.1 Kontext [pro] is the winner for creating a coherent, professional document that looks like a usable 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 Kontext [pro]
- + Excellent photorealistic lighting and texture on the food components.
- + Strong typography integration with a vibrant neon flame effect.
- + High-quality rendering of the background environment and embers.
- − Failed to create an 'exploded' view; the burger components are mostly stacked.
- − Includes duplicate price tags, cluttering the bottom of the composition.
- − Missed the 'starburst' requirement for the price.
Wan 2.5 (Preview)
- + Perfectly captures the 'exploded' request with components dynamically separated in space.
- + Includes the required starburst shape for the price tag.
- + Excellent sense of motion and energetic composition.
- − The 'LIMITED TIME ONLY' text is slightly warped along a curve.
- − The fire effect on the text looks more like dripping honey/syrup than flames.
Verdict: While FLUX.1 Kontext [pro] offers superior photorealistic textures and cleaner typography, Wan 2.5 (Preview) followed the structural prompt instructions much better by actually 'exploding' the burger components. Wan 2.5 (Preview) also correctly included all requested UI elements like the starburst, whereas FLUX.1 Kontext [pro] provided a mostly static stack and redundant price text.
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 Kontext [pro]
- + Excellent text rendering with grainy chalk texture
- + Accurately follows the item descriptions and prices
- + Very realistic board and frame texture
- − Minor spelling errors in the footer text like 'ous' and 'tree'
- − The title is sans-serif instead of the requested elegant cursive
Wan 2.5 (Preview)
- + Beautiful cursive and script-like handwriting style
- + Excellent use of chalk smudges for authentic atmosphere
- + Strong dynamic composition with realistic cafe background
- − Significant text repetition in the third menu item
- − Missing 'Herbs' in the second item and poor alignment of the dollar signs
Verdict: FLUX.1 Kontext [pro] provides much clearer and more accurate text for the menu items themselves, but fails the cursive requirement for the title. Wan 2.5 (Preview) captures a more artistic chalkboard aesthetic with great handwriting style, but suffers from significant logical errors in the text content, such as repeating 'Cookies - $9' twice.
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 Kontext [pro]
- + Perfectly follows the specific instruction of having the horse on top of the astronaut
- + Interprets 'surreal' effectively by blending the horse's legs into the astronaut's suit
- + Highly cinematic lighting and realistic textures
- − The anatomy is intentionally confusing, which might be off-putting to some
- − Includes a small extra rider on the horse which wasn't requested
Wan 2.5 (Preview)
- + Very high resolution and vibrant color palette
- + Good composition with a dynamic action pose
- + Clean rendering of the astronaut suit and horse
- − Fails the negative constraint, depicting a standard astronaut on a horse
- − Relatively generic interpretation of the prompt without the requested surreal reversal
Verdict: FLUX.1 Kontext [pro] successfully followed the difficult conceptual instruction to place the horse on top of the astronaut, resulting in a truly surreal and creative image. Wan 2.5 (Preview) produced a high-quality but traditional image of an astronaut riding a horse, completely ignoring the 'horse on top' requirement.
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 Kontext [pro]
- + Excellent fur texture and lighting on the capybara.
- + Logical placement of the capybara in the driver's seat relative to the steering wheel.
- + The passenger's interaction with her phone is subtle and natural.
- − The passenger is holding her phone to her ear rather than looking at it as requested.
- − The perspective hides most of the NYC background.
Wan 2.5 (Preview)
- + Perfect adherence to the prompt regarding both paws on the steering wheel and the passenger looking at her phone.
- + Excellent composition that captures the vibrant NYC night background while maintaining the interior focus.
- + High level of detail in the taxi accessories like the meter and roof sign.
- − The paws look slightly more like hands/claws than natural capybara feet.
- − The passenger's scale seems slightly small compared to the front seat area.
Verdict: Wan 2.5 (Preview) produced a superior composition that perfectly captured the requested narrative elements, including the passenger's specific action and the iconic NYC backdrop. While FLUX.1 Kontext [pro] has beautiful texture and lighting, it missed the specific detail of the passenger looking at her phone and offered a more limited view of the scene.
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 Kontext [pro]
- + Excellent typography style that matches the vintage gothic aesthetic perfectly.
- + Captures a dark and moody atmosphere with realistic lighting.
- + Highly coherent spider web border that feels integrated into the design.
- − Includes hallucinated text lines that were not in the prompt.
- − The 'dark parchment' feels more like a black background than paper.
Wan 2.5 (Preview)
- + Follows the text instructions perfectly without adding extra hallucinated lines.
- + Features a beautiful thorn and web border that accurately reflects the prompt.
- + The glowing fire inside the jack-o-lantern is visually striking and detailed.
- − The composition feels slightly cluttered due to the large size of the pumpkin and bats.
- − The bright blue moon backdrop slightly clashes with the vintage parchment aesthetic.
Verdict: Wan 2.5 (Preview) is the winner because it adhered perfectly to the requested text without adding the gibberish lines found in the FLUX.1 Kontext [pro] output. While FLUX.1 Kontext [pro] had superior gothic typography, Wan 2.5 (Preview) successfully captured all visual elements like the thorns, bats, and trees in a more balanced, colorful, and accurate manner.
Bald man challenge
Image Editing“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Successfully added thick, high-density hair as requested.
- + Preserved the lighting and background of the original image well.
- + Faces and clothing textures remain sharp and consistent.
- − The facial features were slightly altered, making the man look younger and changing his eye shape.
- − The glasses changed from the source, becoming thicker and darker.
Wan 2.5 (Preview)
- + Excellent preservation of the original facial features and bone structure.
- + Maintained the original glasses and skin texture perfectly.
- + The hair integration feels very natural with a believable hairline.
- − None notable; the edit is highly accurate to the request.
Verdict: Wan 2.5 (Preview) provided a superior edit by successfully adding a full head of hair while perfectly preserving the person's identity, glasses, and facial structure. FLUX.1 Kontext [pro] also succeeded in adding hair, but it noticeably altered the man's facial features and glasses, resulting in a less accurate source preservation.
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 Kontext [pro]
- + Excellent 3D cartoon surface textures on the salmon and rice.
- + Text layout is playful and well-rendered with a subtle 3D effect.
- + Successfully creates a wooden diorama base as requested.
- − The flag icon is stylized as a wavy banner rather than a standard flag icon.
- − The rice grains look slightly like bubbles due to the high gloss.
Wan 2.5 (Preview)
- + Very clean, bold typography that is easy to read.
- + Accurate Japanese flag icon representation.
- + Professional lighting and smooth PBR material finish.
- − The diorama base is a simple cylinder rather than a more interesting isometric miniature base.
- − Text is not centered horizontally relative to the flag icon.
Verdict: Both models followed the instructions well, but FLUX.1 Kontext [pro] captured the 'miniature 3D cartoon' aesthetic more effectively with its wooden diorama base and tactile textures. While Wan 2.5 (Preview) has very clean text and lighting, FLUX.1 Kontext [pro] felt more like a complete isometric scene as requested in the prompt.
Over-the-top cartoon caricature
Editing“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.”
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Strongly preserves the source image's pose and outfit.
- + Effective comic-style illustration technique.
- − Completely ignores the specific prompt instructions regarding dogs, hockey, and the news anchor profession.
Wan 2.5 (Preview)
- + Successfully incorporates all requested thematic elements including the anchor desk, dogs, and hockey gear.
- + High level of creativity in merging the professional and personal interests into a single scene.
- − The facial resemblance to the source image is slightly reduced by the classic caricature style.
Verdict: While FLUX.1 Kontext [pro] creates a high-quality stylized version of the photo, it fails to follow the user's specific editing instructions. Wan 2.5 (Preview) ignores the original composition to fulfill the request, successfully integrating the anchor desk, hockey stick, and dogs as requested.
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 Kontext [pro]
- + Excellent soft lighting and atmosphere with a high level of aesthetic appeal.
- + Beautifully soft fur textures that look natural and high-quality.
- + Coherent and charming character expressions.
- − Failed to include a rabbit, instead showing a second cat with long ears.
- − Characters are sitting rather than chasing and tumbling as requested.
Wan 2.5 (Preview)
- + Successfully included all four distinct species: puppy, tabby kitten, white bunny, and fox kit.
- + Captured the dynamic 'chasing' and 'tumbling' action from the prompt.
- + Vibrant details on the butterflies and flowers with clear dew droplets.
- − The fox's eyes appear somewhat surreal or over-processed.
- − The 'god rays' are a bit harsh and look less natural compared to Model A.
Verdict: While FLUX.1 Kontext [pro] created a more aesthetically pleasing and 'masterpiece' quality image in terms of lighting, it failed significantly on prompt adherence by missing the bunny and the active motion. Wan 2.5 (Preview) correctly identified all four animals and captured the playful energy requested, making it the better response to the specific instructions despite some minor uncanny facial features.
Studio Ghibli Anime Style
Editing“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”
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent watercolor/hand-painted texture consistent with the Studio Ghibli aesthetic.
- + Preserves the specific plaid pattern of the man's shirt very accurately.
- + Accurately translates the facial expressions, including the girlfriend's sharp gaze of disapproval.
- − The colors are slightly more saturated than standard pastel Ghibli palettes.
Wan 2.5 (Preview)
- + Perfectly captures the soft, dreamy Ghibli lighting and color palette.
- + Adds creative thematic elements like floating leaves and sparkles.
- + Excellent character design transformation while maintaining pose and composition.
- − The textures look more like modern digital cel-shading than 'hand-painted' textures.
- − The girlfriend's expression is a bit softened compared to the original's intensity.
Verdict: Both models did an exceptional job of preserving the composition and pose of the 'distracted boyfriend' meme. FLUX.1 Kontext [pro] captures the tactile, hand-painted watercolor texture most associated with Ghibli backgrounds, while Wan 2.5 (Preview) better captures the soft lighting, color grading, and whimsical atmosphere (leaves/sparkles) of the movies.
Golden Hour Stroll
Image Editing“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Successfully adds blowing hair and flying leaves.
- + High degree of source preservation for the background and character.
- + Natural integration of the wind effect on the hair.
- − The leash shape has been slightly altered compared to the original.
- − The woman's left hand position has shifted significantly, now resting on the dog's head.
Wan 2.5 (Preview)
- + Excellent wind effect on hair with realistic motion lines.
- + Excellent source preservation including the woman's gait, hands, and the dog's leash.
- + Leaves feel integrated into the 3D space of the scene.
- − The flying leaves have a slightly cartoonish, neon green appearance.
- − Minor loss of detail in the background trees compared to the original.
Verdict: Both models followed the instructions well, but Wan 2.5 (Preview) is the winner due to superior source preservation, maintaining the woman's posture, hand positions, and the leash geometry perfectly while adding the motion effects. FLUX.1 Kontext [pro] also produced a high-quality image but accidentally modified the woman's left hand and the shape of the leash while applying the edits.
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 Kontext [pro]
- + Excellent typography with proper accented characters
- + Accurate grainy texture that reinforces the vintage feel
- + Clean, minimalist vector execution
- − The steam element is a bit thick and less elegant than Model B
Wan 2.5 (Preview)
- + Elegant illustrative style for the cloche and steam
- + Rich warm color palette with good contrast
- + Sophisticated use of decorative borders and banners
- − Failed to include the accent on 'Caffè'
- − Text alignment within the banner is slightly crowded
Verdict: FLUX.1 Kontext [pro] followed the prompt more precisely, particularly with the correct spelling of 'Caffè' and the requested 'Est. 1720' banner placement. While Wan 2.5 (Preview) produced a more visually ornate and polished illustration, it missed specific typographical details and the 'Est. 1720' isn't explicitly on its own banner as requested.
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 Kontext [pro]
- + Features a bold, large title 'APOLLO 11'
- + Uses a playful, creative layout for the rocket launch
- + Contains detailed lunar surface textures
- − Poor text rendering with several spelling and logic errors
- − Fails to follow the requested sequential steps (1-6)
- − Includes nonsensical visual elements like a planet Saturn with rings near the rocket tip
Wan 2.5 (Preview)
- + Excellent adherence to the requested infographic structure and sequence
- + Highly accurate text rendering for labels and names
- + Stronger 'NASA-inspired' aesthetic with modern, crisp vector iconography
- − Step numbers are missing even though labels are present
- − The 'Collins' icon features a strange helmet/hat compared to the others
Verdict: Wan 2.5 (Preview) is the clear winner as it successfully interprets the request for a sequential infographic with mostly accurate text and logical steps. FLUX.1 Kontext [pro] fails the infographic task by introducing nonsensical scientific elements (like putting Saturn in Earth's atmosphere) and garbling the requested steps.
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