Black Forest Labs' open-weights multimodal flow transformer for in-context image generation and editing, available for non-commercial use with character consistency and style transfer capabilities
Settled by community votes across 17 shared challenges, with an AI judge weighing in on each.
FLUX.1 Kontext [dev]
#54 of 62 in Text-to-Image
FLUX.2 [dev] Flash
#5 of 62 in Text-to-Image
Where the votes landed
FLUX.1 Kontext [dev]
0%
win rate
Ties
0%
FLUX.2 [dev] Flash
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 [dev]
- + Excellent lighting and vibrant colors
- + Very clean glass textures and realistic reflections
- + High level of visual clarity and sharpness
- − The sphere appears slightly large relative to the 'small' descriptor in the prompt
FLUX.2 [dev] Flash
- + Accurately depicts a 'small' sphere as requested
- + More realistic table surface with visible wood grain and knots
- + Naturalistic plant placement and transparency through the glass
- − Image is slightly less sharp than Model A
- − Lighting is a bit flatter and less dynamic
Verdict: Both models followed the complex spatial instructions perfectly. FLUX.1 Kontext [dev] produced a cleaner, more aesthetically vibrant image with striking highlights, while FLUX.2 [dev] Flash captured the scale of the 'small' sphere and the realism of the environment more accurately. FLUX.1 Kontext [dev] is the likely winner for its superior rendering of glass 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 Kontext [dev]
- + Excellent handling of wet pavement reflections and rainy atmosphere.
- + High visual clarity and clean rendering of the subject's face.
- + Good implementation of shallow depth of field with the background bokeh.
- − The subject is simply sitting on/posing with the bike rather than repairing it.
- − The composition feels very staged and centered, missing the 'candid' and 'imperfect framing' requested.
- − Cars in the background are static with no motion blur.
FLUX.2 [dev] Flash
- + Perfectly captures the 'repairing' aspect of the prompt with tools and an active pose.
- + Accurately represents 'imperfect framing' and 'candid' street photography style.
- + Includes realistic motion blur on passing cars as requested.
- − The bike's physical structure has some illogical geometry (e.g., the front fork and handlebars).
- − Lower overall resolution and slightly muddier textures compared to Model A.
Verdict: Model B (FLUX.2 [dev] Flash) is the clear winner for prompt adherence, accurately depicting the man in the act of repairing the bike with motion-blurred cars and an authentic candid feel. Model A (FLUX.1 Kontext [dev]) produced a more polished image with better lighting, but it failed to include the primary action of repairing and ignored the specific 'motion blur' and 'candid/imperfect' framing instructions.
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 [dev]
- + Excellent use of warm lighting and dramatic shadows
- + High-quality skin texture and lifelike eyes
- + Strong interpretation of 'ornate engraved plate armor'
- − Missed the request for braided hair with beads
- − The background is very dark, losing some sense of environment
FLUX.2 [dev] Flash
- + Perfect adherence to the braiding and beads requirement
- + Superior rendering of leather straps, buckles, and cloth underlayers
- + Captures the 'battle-worn' aspect with more realistic scars and dirt
- − Lighting is somewhat flat compared to the requested 'warm torchlight'
- − Background torches feel a bit repetitive and symmetrical
Verdict: While FLUX.1 Kontext [dev] creates a more cinematically lit portrait with impressive facial detail, it fails to include the specific hair requirements. FLUX.2 [dev] Flash captures every detail of the prompt, including the complex braids, beads, and varied textures of the armor and leather, making it the more accurate and technically complete image.
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 [dev]
- + Features a bold, high-contrast editorial layout
- + High-quality, vibrant food photography with good clarity
- + Strong adherence to the minimalist grid concept
- − Text is largely nonsensical and features glitches/artifacts
- − The pizza section requested is not clearly identifiable in the food photos
FLUX.2 [dev] Flash
- + Excellent adherence to the menu structure with clear Appetizer, Pizza, and Main sections
- + Includes 'vibrant accents' with colored frame borders as requested
- + Legible prices and cleaner text rendering than the competitor
- − Repetitive food photos (mostly pizzas) even in the non-pizza sections
- − Slightly less 'minimalist' than requested due to the busy header and footer
Verdict: FLUX.2 [dev] Flash is the preferred choice because it successfully followed the structural requirements of the prompt, including specific menu sections for appetizers, pizza, and mains. While FLUX.1 Kontext [dev] had more variety in its food photography and a cleaner aesthetic, its failure to organize the text into logical sections and its garbled character rendering made it less effective as a menu design.
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 [dev]
- + Excellent typography rendering with clean, professional font styles
- + Strong composition with a focus on a high-detail, appetizing burger
- + Great use of lighting and color, specially the glowing embers at the base
- − Failed to follow the instruction for an 'exploded' burger, showing an assembled one instead
- − Minor typo in the secondary text: 'LNHLY' instead of 'ONLY'
FLUX.2 [dev] Flash
- + Perfect adherence to the 'exploded' burger instruction with suspended components
- + Text perfectly matches the 'fiery, glowing effect' requested in the prompt
- + Accurate spelling on all requested text elements
- − The 'starburst' shape for the price is a bit irregular
- − The sauce droplets look slightly repetitive in their placement
Verdict: FLUX.2 [dev] Flash is the clear winner as it successfully rendered the exploded view of the burger, whereas FLUX.1 Kontext [dev] generated a standard assembled burger. Additionally, FLUX.2 [dev] Flash followed the stylistic text requirements for a fiery effect and avoided the spelling error found in FLUX.1 Kontext [dev].
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 [dev]
- + Features a bold, high-contrast chalkboard aesthetic.
- + Captures the chalkboard texture well in the 'TODAY SPECIALS' heading.
- − Numerous spelling errors including 'Risoktso', 'Mashroom', and 'Octpus'.
- − Failed to render the date correctly, using garbled 'H1VRIII' characters.
- − Significant repetition and layout clutter around the pricing and herbal description.
FLUX.2 [dev] Flash
- + Excellent spelling accuracy for nearly all requested text.
- + Highly realistic chalk texture with smudges and natural variations.
- + Superior composition that effectively captures the 'cozy café' atmosphere in the background.
- − Repeated the price and the word 'Cookies' on a second line for the final item.
- − The cursive for the title is very subtle rather than 'elegant cursive'.
Verdict: FLUX.2 [dev] Flash is the clear winner as it successfully rendered the complex text and specific date without the significant spelling and repetition errors found in FLUX.1 Kontext [dev]. Additionally, FLUX.2 contextualized the board in a realistic café environment, whereas FLUX.1 provided a flat, isolated board with jumbled characters.
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 [dev]
- + Excellent adherence to the complex spatial instruction of 'horse on top'
- + Clean, minimalist composition that emphasizes the surreal nature of the prompt
- + High clarity in the rendering of the astronaut's face and suit
- − The 'riding' aspect is slightly ambiguous as the horse is floating behind/above rather than sitting on the astronaut
- − Background is relatively empty compared to the cinematic request
FLUX.2 [dev] Flash
- + Beautiful cinematic lighting and detailed space background
- + High level of texture on the horse's coat and mane
- − Completely failed the negative constraint to have the horse on top
- − Produces a cliché AI trope instead of following the specific surreal instruction
Verdict: FLUX.1 Kontext [dev] is the clear winner as it is the only model that successfully followed the difficult prompt logic of placing the horse on top of the astronaut. While FLUX.2 [dev] Flash produced a visually stunning image with better cinematic atmosphere, it completely ignored the core instruction, resulting in a standard, unoriginal interpretation.
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 [dev]
- + High resolution with excellent texture on the capybara's fur
- + Good dramatic lighting that creates a photorealistic mood
- + Excellent rendering of the human passenger's hands and phone
- − The capybara only has one paw on the steering wheel, failing the 'both front paws' instruction
- − The capybara's snout is shortened, making it look more like a groundhog or beaver
FLUX.2 [dev] Flash
- + Perfectly adheres to the prompt by showing both paws on the steering wheel
- + Captures the classic capybara facial structure much more accurately
- + Better cinematic composition showing the 'TX' taxi sign and blurred Manhattan lights
- − The human passenger's hands and phone are slightly blurry and less defined than in the other image
- − Minor artifacts where the paws meet the steering wheel
Verdict: While FLUX.1 Kontext [dev] has slightly more crisp textures, FLUX.2 [dev] Flash is the clear winner for its superior prompt adherence and character accuracy. FLUX.2 correctly depicted both paws on the wheel and maintained the distinct long snout of a capybara, whereas FLUX.1 failed the paw instruction and misinterpreted the animal's features.
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 [dev]
- + Strong, legible main title typography
- + Bold central jack-o-lantern illustration
- + Accurate date and time placement
- − Text in the scroll banner and location is garbled and illegible
- − Background lacks the requested moody night sky and parchment texture
- − Composition feels a bit flat and graphic rather than cinematic
FLUX.2 [dev] Flash
- + Excellent atmosphere with moody lighting and parchment texture
- + Perfectly legible text throughout the entire invitation
- + Complex and detailed border featuring webs and thorns as requested
- − Slightly odd characters appear next to the 'Time' section
- − Main title font is slightly less 'heavy' than model A
Verdict: FLUX.2 [dev] Flash significantly outperformed FLUX.1 Kontext by following all aspects of the prompt, including the complex border details and moody atmospheric lighting. Most importantly, FLUX.2 [dev] Flash maintained high text accuracy across all sections, whereas FLUX.1 Kontext struggled with the banner and location text.
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 [dev]
- + Strong prompt adherence regarding thickness and density.
- + Follows the request for a clean, natural hairline.
- − Significantly alters facial features, making the man look younger and like a different person.
- − The hair texture and lighting on the head appear overly smooth and AI-generated compared to the rugged source.
FLUX.2 [dev] Flash
- + Excellent preservation of original facial features, glasses, and skin texture.
- + Seamlessly blends the new hair with the existing beard and background or environment.
- − The 'afro' style may be an unexpected interpretation of 'natural hair' for this specific person's features.
- − Slight blurring where the hair meets the top of the forehead.
Verdict: FLUX.2 [dev] Flash is the clear winner for its superior source preservation, keeping the man's face and original details almost perfectly intact while adding the requested hair. FLUX.1 Kontext [dev] fails as an edit model because it regenerates the entire face, resulting in a person that no longer resembles the source image.
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 [dev]
- + Excellent adherence to the 'cartoon scene' and 'soft refined textures' prompt description
- + Bold and readable text typography
- + Clean, minimalist composition that feels professional and intentional
- − The flag icon is stylized incorrectly, appearing as an abstract shape rather than the Japanese flag
- − The sushi anatomy is slightly nonsensical with the rice placement inside the wrap
FLUX.2 [dev] Flash
- + Highly accurate representation of the Japanese flag icon
- + Rich, detailed textures on the fish and wasabi that still maintain a miniature feel
- + Superior isometric layout with 45-degree top-down perspective
- − The text is a bit small and simple compared to the 'bold' request
- − The sushi rolls are slightly crowded on the plate
Verdict: FLUX.2 [dev] Flash is the winner as it accurately rendered the flag and provided a much more sophisticated 'miniature diorama' feel with realistic materials. While FLUX.1 Kontext [dev] had a nice cartoon aesthetic, its failure to render a recognizable flag and its strange sushi structure made it less effective overall.
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 [dev]
- + Captures the user's likeness in a comic-book style.
- + Follows the request for a dog and a TV anchor profession.
- + Maintains the composition of the original selfie.
- − Completely misses the 'hockey' requirement.
- − Visual quality is a bit flat and the dog character is poorly integrated.
- − Text 'JOB' is a bit literal and uninspired.
FLUX.2 [dev] Flash
- + Highly vibrant caricature style with excellent 'big head' exaggeration.
- + Successfully incorporates all three elements: TV anchor desk, multiple dogs, and a detailed hockey rink setting.
- + Excellent preservation of the subject's facial features and expression while translating to an artistic style.
- − The 'TV SHOW' text is a bit generic.
- − Minor artifacts in dog paws/hockey sticks upon close inspection.
Verdict: FLUX.2 [dev] Flash is the clear winner as it successfully incorporated every element of the prompt (TV anchor, dogs, and hockey) into a cohesive and humorous caricature. FLUX.1 Kontext [dev] failed to include the hockey theme and had a much simpler, less professional-looking artistic style compared to the high-detail work of FLUX.2.
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 [dev]
- + Excellent sense of motion and playfulness with the 'tumbling' action
- + Clean, vibrant aesthetics with warm lighting effects
- + High clarity on the foreground flowers and grass
- − Failed to include the baby bunny and red fox kit requested in the prompt
- − Anatomical issues with the rightmost animal, which appears to be a hybrid of a kitten and puppy
FLUX.2 [dev] Flash
- + Successfully included all requested animals: golden retriever, tabby kitten, bunnies, and red fox kits
- + Beautiful rendering of dew sparkles on the grass as requested
- + Excellent execution of 'god rays' and sunrise lighting
- − The composition is a static group portrait rather than the 'chasing and tumbling' action requested
- − Some minor fur clipping where the animals overlap in the huddle
Verdict: While FLUX.1 Kontext [dev] captures the energetic 'tumbling' vibe of the prompt better, it fails significantly on prompt adherence by omitting half of the requested animals. FLUX.2 [dev] Flash is much more faithful to the specific animal types and environmental details like dew sparkles, making it the superior choice for following complex instructions despite the more static composition.
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 [dev]
- + Excellent preservation of the original city background elements.
- + Accurately recreates the character poses and clothing patterns.
- + Clean, modern anime aesthetic that mimics high-quality cel shading.
- − Fails to capture the soft pastel and hand-painted texture requirement.
- − The lighting feels flat and digital rather than nostalgic or dreamy.
FLUX.2 [dev] Flash
- + Perfectly captures the Studio Ghibli 'dreamy' aesthetic with soft pastels and watercolor textures.
- + Successfully integrates a nostalgic, painterly mood into the image.
- + Maintains the core character composition while enhancing the artistic style.
- − Replaces the urban street background with a field of flowers, losing the original context.
- − The man's facial expression is slightly less expressive than the original.
Verdict: FLUX.1 Kontext [dev] did a better job of preserving the source image context and layout, but it looks like a standard modern anime rather than the specific Studio Ghibli style requested. FLUX.2 [dev] Flash perfectly captured the Ghibli artistic requirements—including soft textures and warm lighting—even though it altered the background significantly. FLUX.2 [dev] Flash is the winner for better adhering to the specific stylistic keywords of the prompt.
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 [dev]
- + Successfully added wind effects to the hair in a realistic manner.
- + Preserved the identity and facial features of the woman and dog almost perfectly.
- − Included very few leaves, failing to fully capture the requested energy.
- − The overall image feels static despite the hair movement.
FLUX.2 [dev] Flash
- + Strong adherence to the 'flying leaves' instruction with a high volume of particles.
- + Dramatic and energetic hair movement that effectively conveys motion.
- + Excellent preservation of the source image backgrounds and subjects.
- − Some leaves in the foreground appear slightly blurry or lack integration with the lighting.
Verdict: FLUX.2 [dev] Flash followed the instructions much more effectively by adding a significant amount of flying leaves and a more dramatic wind effect on the hair, creating a truly energetic feel. FLUX.1 Kontext [dev] was much more subtle, only slightly modifying the hair and adding a few tiny leaves, which failed to transform the energy of the photo as requested.
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 [dev]
- + Strong minimalist vector aesthetic
- + Perfectly legible and bold typography
- + Clean execution of the cloche dome concept
- − Missed the request for a banner element
- − Design feels slightly modern rather than vintage
- − Steam element is very abstract
FLUX.2 [dev] Flash
- + Excellent adherence to all prompt elements including the banner and steam
- + Sophisticated vintage texture and color palette
- + Beautiful classic typography that fits the restaurant theme
- − Slightly less 'minimalist' than a standard flat vector logo
Verdict: Both models performed well, but FLUX.2 [dev] Flash is the clear winner for its superior adherence to the specific composition requested, including the banner and textured background. While FLUX.1 Kontext [dev] produced a clean logo, it ignored the banner requirement and the vintage texture was almost non-existent compared to the rich, aged feel of the winner.
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 [dev]
- + Follows the requested NASA-inspired navy and red color palette well.
- + Maintains a consistent, minimalist vector art style across all icons.
- − Very poor text rendering with many spelling errors like 'APOLO' and illegible gibberish.
- − The layout of the steps is confusing and non-linear, making it fail as an infographic.
FLUX.2 [dev] Flash
- + Excellent layout that logically follows the mission steps from top to bottom.
- + Superior text rendering for names and mission phases.
- + Highly detailed illustrations that still fit the flat-vector aesthetic.
- − Minor text artifacts on words like 'LAUNCH'.
- − Repeats the 'LUNAR ORBIT' label twice in the central section.
Verdict: FLUX.2 [dev] Flash is the clear winner as it successfully creates a functional infographic with a logical flow, legible text, and high-quality illustrations that match the prompt. FLUX.1 Kontext [dev] fails on basic spelling ('APOLO') and produces icons that are abstract and difficult to interpret as specific mission phases.
Explore each model
Fast distilled version of Black Forest Labs' FLUX.2 [dev] optimized for speed and cost efficiency.