Black Forest Labs' premium multimodal flow transformer with greatly improved prompt adherence and typography generation for in-context image generation and editing without compromise on speed
Settled by community votes across 17 shared challenges, with an AI judge weighing in on each.
FLUX.1 Kontext [max]
#23 of 62 in Text-to-Image
FLUX.2 [dev]
#19 of 62 in Text-to-Image
Where the votes landed
FLUX.1 Kontext [max]
100.0%
win rate
Ties
0.0%
FLUX.2 [dev]
0.0%
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 [max]
- + Excellent handling of caustics and light refraction on the wooden surface.
- + The scale of the blue sphere is well-proportioned to the cube.
- + High level of texture detail on the book cover and the sphere.
- − The plant is largely behind a solid dark background rather than being clearly visible through the glass cube as requested.
- − The cube structure has some slightly inconsistent edge thickness.
FLUX.2 [dev]
- + Perfect adherence to the request of showing the plant through the glass.
- + Includes a visible window in the composition to justify the light source.
- + Clean and realistic glass geometry with logical reflections.
- − The sphere appears slightly large for the described 'small sphere'.
- − The lighting on the table surface is more diffuse and lacks the sharp caustic interest found in the other image.
Verdict: FLUX.1 Kontext [max] features more dramatic lighting and impressive surface textures, but FLUX.2 [dev] follows the spatial instructions of the prompt more accurately by placing the plant directly behind the cube so it is visible through the glass. FLUX.2 [dev] also provides a more coherent scene by including the window mentioned in the lighting prompt.
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 [max]
- + Excellent depiction of rain streaks and realistic pavement reflections.
- + Accurate anatomical handling of the man's hands and grip on the bicycle chain.
- + Compelling cinematic atmosphere with lighting that feels grounded and natural.
- − Missed the request for motion blur from passing cars, as the background vehicles are static.
- − The framing feels a bit too composed and center-balanced for the 'imperfect framing' prompt.
FLUX.2 [dev]
- + Perfect execution of the requested motion blur from passing cars in the background.
- + Highly realistic skin textures and flyaway hairs on the subject.
- + Better adherence to the 'imperfect framing' prompt with a more candid, top-down perspective.
- − Physical bike architecture is messy, with brake cables and handlebars looking nonsensical.
- − The man's hands have significant anatomical errors, including an extra thumb/finger merging into the metal.
- − Severe perspective distortion on the bicycle frames.
Verdict: FLUX.2 [dev] significantly better captures the 'motion blur' and 'candid' aspects of the prompt, but suffers from severe structural AI artifacts in the bicycle and hands. FLUX.1 Kontext [max] produces a much more coherent and high-quality image with correct anatomy and mechanical details, though it fails to include the requested motion blur. FLUX.1 Kontext [max] is the winner for its overall technical reliability and realistic lighting.
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 [max]
- + Exceptional realistic skin and facial hair textures.
- + Intricate and high-fidelity engraving details on the armor plates.
- + Stronger implementation of warm torchlight reflections on the metal.
- − Failed to include the requested beads in the braided hair.
- − Scars are very faint and less noticeable compared to the prompt requirements.
FLUX.2 [dev]
- + Successfully included small colored beads in the braids.
- + More accurate representation of a 'battle-worn' appearance with visible scars and dirt.
- + Good depth of field with visible torchlight source.
- − Skin texture appears slightly smoother and less 'lifelike' than the competitor.
- − Armor engravings are less detailed and look slightly more generic.
Verdict: FLUX.1 Kontext [max] produces a much more realistic and high-textured portrait, though it misses the specific detail of beads in the hair. FLUX.2 [dev] adheres more closely to all prompt instructions, including the beads and more prominent scarring, but falls short on the hyper-realistic skin and metal textures found in the other model.
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 [max]
- + The food photos have a consistent, high-key studio style that feels professional.
- + The top-down grid layout is very clean and symmetrical.
- + The logo and main title area are well-integrated into the design.
- − Failed to provide distinct sections for appetizers, pizza, and mains as requested.
- − The body text is largely gibberish and includes script fonts that contradict the 'bold sans-serif' prompt.
- − Over-focused on pizza, lacking variety in the food items.
FLUX.2 [dev]
- + Successfully included all three requested sections: Appetizers, Pizza, and Mains.
- + The layout is more dynamic and closer to a functional double-page menu.
- + Strictly adheres to the bold sans-serif font requirement for headings.
- − The food photos are slightly inconsistent in lighting and plate styling.
- − The price digits are unrealistic (e.g., $990 for a pizza).
- − Text rendering within the menu items is messy compared to the headings.
Verdict: FLUX.2 [dev] is the clear winner as it accurately followed the prompt's structural requirements for specific sections (Appetizers, Pizza, Mains) and font styles. FLUX.1 Kontext [max] produced a cleaner single-page aesthetic but ignored the content variety requested, filling the page almost entirely with pizza images.
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 [max]
- + Excellent photorealistic texture on the meat patty and bun
- + Clear, legible typography for all text components
- + Effective background atmosphere with detailed embers and sparks
- − The burger is not truly 'exploded' as much as it is a stack with random food pieces floating beside it
- − Failed to place the price in a starburst, just placed it at the bottom
- − The burger composition feels slightly static rather than high-motion
FLUX.2 [dev]
- + Perfect adherence to the 'exploded' request with vertical separation of all layers
- + Successfully followed the instruction to place the price in a fiery starburst
- + Superior sense of motion with sauce droplets flying outwards
- − Typography has a slightly more artificial, digital glow look compared to the cohesive lighting in Image A
- − The lettuce and bottom bun look slightly less realistic than those in Image A
Verdict: FLUX.2 [dev] followed every instruction of the prompt much more accurately, specifically the 'exploded' layout and the starburst element for the price. While FLUX.1 Kontext [max] has slightly better photorealistic textures on the burger itself, it failed to separate the layers vertically as requested, making it feel less dynamic.
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 [max]
- + Excellent text legibility and alignment.
- + Accurately completes the 'Brown Butter' item prompt.
- + Effective use of negative space on the chalkboard.
- − The title is in print-style block letters rather than the requested elegant cursive.
- − The chalk texture is very clean, appearing slightly more like a digital chalk-font than a natural rubbing.
FLUX.2 [dev]
- + Successfully captures the elegant cursive/slant requested in the prompt.
- + Superior chalk texture with realistic smudges and varying pressure.
- + More natural variations in letter size and baseline.
- − The text 'Grilled Octopus with Lemon & Herbs' has some messy overlapping and artifacts near the word 'with'.
- − The title cursive is a bit shaky, affecting legibility of 'Specials'.
Verdict: FLUX.2 [dev] followed the stylistic instructions much better, delivering the requested cursive title and a very realistic chalk texture that feels handwritten. FLUX.1 Kontext [max] produced much cleaner text, but it failed to use cursive for the title and the result looks slightly more like a digital font overlay rather than a real chalkboard.
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 [max]
- + Excellent lighting and cinematic color grading.
- + Crisp details on the space suit and horse's fur.
- + Cleaner background with a more realistic atmospheric glow.
- − Prompt adherence failure regarding the positioning instruction.
- − The horse's front legs appear slightly stiff or anatomically awkward.
FLUX.2 [dev]
- + Dynamic pose with a sense of motion in the horse's gallop.
- + Good inclusion of a detailed galaxy and planetary surface in the background.
- + Strong textural details on the horse's mane and tack.
- − Prompt adherence failure regarding the positioning instruction.
- − Odd shadow placement on the clouds below that doesn't match the light source.
Verdict: Both FLUX.1 Kontext [max] and FLUX.2 [dev] failed to follow the specific surreal instruction for the horse to be 'on top' of the astronaut, instead providing standard 'astronaut on horse' images. FLUX.1 Kontext [max] is the preferred image due to its superior lighting, cleaner composition, and more professional cinematic finish compared to FLUX.2 [dev].
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 [max]
- + Excellent fur texture rendering on the capybara
- + High quality cinematic lighting
- + Creative interpretation of a modern driver's hat
- − The passenger is on a call rather than looking at her phone as requested
- − Only one paw is visible on the steering wheel instead of both
- − Composition cuts off much of the vehicle interior
FLUX.2 [dev]
- + Perfect adherence to the phone-usage and bored expression details
- + Accurately depicts both paws on the steering wheel
- + Stronger interior taxi composition providing more environmental context
- − The passenger's hands and the capybara's paws have some slight structural uncanny valley effects
- − The 'T' logo on the hat is a bit generic compared to the 'TAXI' text in image A
Verdict: While FLUX.1 Kontext [max] has superior texture quality and lighting, FLUX.2 [dev] followed the complex prompt instructions much more accurately. FLUX.2 [dev] correctly placed both paws on the wheel and depicted the passenger looking at her phone with a bored expression, whereas FLUX.1 Kontext [max] had the passenger on a phone call.
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 [max]
- + Excellent atmospheric lighting around the jack-o-lantern.
- + Intricate and high-quality gothic border design.
- + Strong cinematic mood with darker color gradients.
- − Redundant text displaying 'The Arches, NYC' twice at the bottom.
- − The date contains unnecessary commas instead of dots as requested.
FLUX.2 [dev]
- + Perfect text accuracy for all custom details including the date and location.
- + Clearer scroll banner implementation for the secondary text.
- + Very well-defined thorny border and spiderwebs.
- − The lighting is a bit flatter and less 'cinematic' than the prompt suggests.
- − The jack-o-lantern rendering is slightly more illustrative and less atmospheric.
Verdict: Both models followed the complex prompt very well, but FLUX.2 [dev] is the winner due to superior text accuracy and layout organization, avoiding the duplication errors found in FLUX.1 Kontext [max]. While FLUX.1 Kontext [max] had a more moody and cinematic aesthetic, the incorrect punctuation in the date and the repetitive location line made it less functional as an invitation.
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 [max]
- + Expertly blends the new hair onto the original skull shape.
- + Preserves the original facial orientation and features perfectly.
- + High-quality hair texture that matches the lighting of the scene.
- − The hairline is slightly too sharp and low on the forehead.
FLUX.2 [dev]
- + Impressive preservation of the original background and clothing.
- + Clearly interprets 'thick' with a very voluminous style.
- − The hair texture doesn't match the beard texture well.
- − The transition between the hair and the forehead/temples looks unnatural and poorly blended.
- − The volume feels disconnected from the person's head shape.
Verdict: FLUX.1 Kontext [max] provides a much more natural and believable hair transplant, integrating the new strands seamlessly with the original scalp and existing beard. While FLUX.2 [dev] follows the 'thick' instruction by providing an afro, the blending around the hairline is poor and the style feels like a separate layer rather than part of the original man.
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 [max]
- + Excellent 3D cartoon styling with vibrant, appealing fish textures.
- + Legible typography that matches the bubble-style aesthetic.
- + Very clean and balanced composition for the diorama base.
- − Failed to include the requested flag icon.
- − The text is brown instead of the more standard black or white for such icons.
FLUX.2 [dev]
- + Successfully included all prompt elements including the small flag icon.
- + Higher variety of sushi types (nigiri and maki) presented in the scene.
- + Clean, professional typography.
- − The lighting is a bit more muted compared to the 'gentle' yet clear lighting requested.
- − Small shadow artifacts on the top surface of the diorama base.
Verdict: FLUX.2 [dev] is the winner as it adhered to every specific detail of the prompt, including the flag icon which FLUX.1 Kontext [max] omitted. While FLUX.1 Kontext [max] had slightly more vibrant textures on the fish, FLUX.2 [dev] provided a more complete and accurate interpretation of the requested diorama.
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 [max]
- + Excellent preservation of the subject's clothing and basic features in a cartoon style.
- + Clear incorporation of all elements: news desk, dog, and hockey stick.
- + Bold, clean illustration style with high legibility.
- − The 'caricature' aspect is less defined, looking more like a standard avatar than an exaggerated caricature.
- − Added glasses that were not in the source image.
FLUX.2 [dev]
- + Perfectly captures the 'caricature' style with an exaggerated head and stylized features.
- + Highly creative integration, placing the news desk directly on a hockey rink with dogs in jerseys.
- + Excellent facial likeness despite the stylistic exaggeration.
- − Text on the news desk is nonsensical gibberish.
- − The secondary human character in the top left is slightly distracting from the main subject.
Verdict: FLUX.2 [dev] followed the 'caricature' instruction much better than FLUX.1 Kontext [max], providing the classic big-head style and a more imaginative scene layout. FLUX.2 [dev] also achieved a stronger facial resemblance to the source image while successfully integrating the hockey and dog elements in a humorous way. FLUX.1 Kontext [max] is a high-quality illustration, but it feels more like a generic cartoon character than a personalized 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.”
AI Judge Analysis
FLUX.1 Kontext [max]
- + Excellent adherence to the 'fluffy' and 'ultra-detailed soft fur' part of the prompt
- + Captures a very bright, joyful, and magical atmosphere with vibrant colors
- + Stronger execution of 'god rays' and sun sparkles across the whole frame
- − The bunny appears slightly stylized/cartoonish compared to the other animals
- − The composition feels a bit crowded with many butterflies overlapping the subjects
FLUX.2 [dev]
- + More realistic fur textures and anatomical proportions for the animals
- + The lighting feels more natural and grounded while still being warm
- + Better clarity and detail on the butterflies
- − Added an extra fifth animal (a second bunny) which was not requested in the prompt
- − The lower half of the image is somewhat dark, losing some of the 'lush wildflower' detail
Verdict: FLUX.1 Kontext [max] creates a more whimsical and magical scene that perfectly matches the 'wholesome' and 'fluffy' keywords, though some animals look a bit less realistic. FLUX.2 [dev] provides higher photographic fidelity and realistic lighting but fails the count check by adding an extra bunny. FLUX.1 Kontext [max] is preferred for better capturing the specific aesthetic 'vibe' and requested animal count.
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 [max]
- + Excellent preservation of the source image's composition and urban background
- + Perfectly captures the Studio Ghibli character aesthetic with clean line work and expressive faces
- + Maintains the specific clothing colors and patterns from the original meme
- − The textures are a bit flatter and less 'painterly' than the Ghibli background style
FLUX.2 [dev]
- + Beautiful use of soft pastel colors and dreamy, hand-painted lighting
- + Successfully captures the 'nostalgic mood' requested in the prompt
- + Artistic texture reflects watercolor aesthetic well
- − Replaces the urban street background with a flower field, failing to preserve that part of the source image
- − Character likeness and expressions are slightly less accurate to the original meme than Model A
- − The hand of the woman in the foreground is missing
Verdict: FLUX.1 Kontext [max] is the winner as it perfectly translates the 'distracted boyfriend' meme into a Ghibli aesthetic while maintaining the recognizable urban setting and character expressions. While FLUX.2 [dev] produces a very beautiful and painterly image, it takes too much creative liberty by removing the street setting and replacing it with a generic flower field, which detracts from the context of the source image.
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 [max]
- + Excellent preservation of the woman's face and original features.
- + Changes the pose of the arms and dog to create a more 'energetic' walking motion as requested.
- + Realistic hair movement that feels natural to the scene.
- − The leash logic becomes a bit messy near the dog's collar.
- − The woman's left hand is poorly rendered with too many fingers.
FLUX.2 [dev]
- + Highly dynamic hair effect that perfectly captures the feeling of wind.
- + Added a substantial number of detailed leaves to enhance the motion theme.
- + Maintained the original pose and background with very high fidelity.
- − The hair feels slightly 'pasted on' at the edges compared to the original lighting.
- − A leaf is awkwardly floating directly on the woman's lower leg/denim.
Verdict: FLUX.1 Kontext [max] took a more transformative approach by changing the woman's posture to a more active walk, but suffered from significant anatomy issues on the hands. FLUX.2 [dev] followed the edit instructions more effectively by adding dramatic wind effects to the hair and a flurry of leaves while maintaining the high quality and anatomical correctness of the original image.
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 [max]
- + Excellent typography with a bold, classic serif font.
- + Perfect alignment of elements and superior layout balance.
- + High-contrast textures that give it a premium printed feel.
- − The accent on the 'Ê' is slightly oversized compared to standard typography.
FLUX.2 [dev]
- + Clean vector-style execution.
- + Good use of the banner for both the name and the date.
- − The 'CAF' in the banner is slightly warped and not as legible as Model A.
- − The background is a bit too yellow rather than cream.
- − Small stray graphical artifacts below the main banner.
Verdict: FLUX.1 Kontext [max] produced a much more professional and aesthetically pleasing logo with superior typography and texture. FLUX.2 [dev] follows the prompt well, but the layout is less balanced and the text within the banner is noticeably less refined than the text in FLUX.1 Kontext [max].
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 [max]
- + Excellent typography and clean heading
- + High-quality vector silhouettes and icon of the Lunar Module
- + Successfully includes names associated with the mission
- − Confused flowchart logic where Earth is smaller than the Moon
- − Only includes 4 of the requested 6 steps
- − Nonsense filler text present in the middle of the diagram
FLUX.2 [dev]
- + Follows the requested dark navy and muted red palette perfectly
- + Includes all specific icons requested in a grid format
- + Superior adherence to the chronological steps outlined in the prompt
- − Several labels feature garbled or misspelled text
- − Icons are somewhat repetitive and crowded
- − Literal interpretation of 'Saturn V icon' resulted in it being labeled 'SATURN ICON'
Verdict: FLUX.2 [dev] is the winner for its superior prompt adherence, successfully capturing all six steps of the mission and the specific NASA-inspired color palette. While FLUX.1 Kontext [max] has cleaner graphic design and better typography, it failed to provide the full sequence and has a confusing visual hierarchy regarding the Earth and Moon.
Explore each model
Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment