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] Turbo
#3 of 62 in Text-to-Image
Where the votes landed
FLUX.1 Kontext [max]
50.0%
win rate
Ties
0.0%
FLUX.2 [dev] Turbo
50.0%
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 light and caustic reflections on the table.
- + High cinematic quality with rich textures on the sphere and wooden table.
- + Perfect logical placement of all requested elements.
- − The plant is slightly less 'behind' the cube compared to Model B, appearing more to the side.
FLUX.2 [dev] Turbo
- + Successfully places the plant directly behind the cube as requested.
- + Realistic glass thickness and texture on the cube.
- + Accurate interpretation of the soft window light from the left.
- − The blue sphere is merged with a vertical refraction line, making it look slightly distorted.
- − The table texture is a bit more generic compared to the high-contrast grain in Model A.
Verdict: Both models followed the complex spatial prompt perfectly. FLUX.1 Kontext [max] produced a more visually striking image with superior lighting effects and richer textures, while FLUX.2 [dev] Turbo did a slightly better job positioning the plant specifically 'behind' the glass, despite some strange refraction artifacts in the center.
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 atmosphere
- + Strong adherence to 'light rain' and 'cinematic' lighting
- + Realistic skin texture and hand positioning
- − The man's ethnicity is somewhat ambiguous and doesn't look distinctly Japanese
- − Missing the requested motion blur from passing cars
- − The bicycle geometry is slightly warped where the chain meets the frame
FLUX.2 [dev] Turbo
- + Includes the motion blur from passing cars as requested
- + More accurate representation of an elderly Japanese man
- + Captures the 'imperfect framing' and 'candid' feel perfectly
- − The front wheel of the bicycle is clipping awkwardly through the pavement
- − Visual quality is slightly lower with more noisy artifacts compared to Model A
- − Rain is less visible than in Model A
Verdict: FLUX.2 [dev] Turbo captures more specific prompt elements like the motion blur of cars and the specific ethnicity of the subject, though it suffers from a significant clipping artifact where the bike wheel enters the ground. FLUX.1 Kontext [max] produces a higher-quality, more cinematic image with better rain effects, but fails to include the requested car motion and the subject's appearance is less localized to the requested setting. FLUX.2 [dev] Turbo is the winner for better overall prompt adherence and 'candid' feel despite the technical clipping error.
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]
- + Excellent depiction of warm torchlight reflecting off the polished metal surfaces.
- + Extremely high skin texture detail with visible pores and realistic moisture/sweat.
- + Beautifully intricate floral engravings on the armor plate.
- − Missed the request for small beads in the braided hair.
- − The scars are very faint, leaning more towards a clean look than 'battle-worn'.
FLUX.2 [dev] Turbo
- + Perfect adherence to specific details like the small beads in the hair and multiple scars.
- + Outstanding texture on leather straps and the cloth underlayer as requested.
- + Stronger storytelling with the background torch and additional elements like the chainmail.
- − The lighting feels slightly flatter and less dramatic than image A.
- − The bokeh sparks appear a bit more repetitive/uniform across the frame.
Verdict: While FLUX.1 Kontext [max] produces a more striking high-end cinematic portrait with superior lighting and engraving detail, FLUX.2 [dev] Turbo followed the prompt instructions more closely regarding specific features like hair beads and battle-worn scars. Image B's inclusion of detailed leatherwork and cloth gives it a slight edge in terms of total prompt adherence.
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]
- + Excellent variety of high-quality food photography
- + Cleaner overall layout with consistent white space
- + Professional use of a hero logo at the top
- − Included script/italic fonts which goes against the 'bold sans-serif' requirement
- − Text content is mostly nonsensical gibberish
FLUX.2 [dev] Turbo
- + Strictly followed the 'bold sans-serif fonts' and 'sections' requirements
- + Better adherence to the requested grid layout for food photos
- + Includes clear pricing and legible category headers like APPETIZERS and MAINS
- − Bottom color bar is a bit distracting for a minimalist design
- − The pricing numbers are illogical (e.g., $260 for pizza)
Verdict: FLUX.2 [dev] Turbo followed the prompt's specific instructions for sections and sans-serif typography much better than FLUX.1 Kontext [max]. While FLUX.1 Kontext [max] has a slightly cleaner aesthetic, it failed to provide the requested category sections and used script fonts which deviated from the brief.
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, high-contrast typography
- + Dynamic debris and embers enhance the sense of motion
- − Failed to render the price inside a starburst
- − The burger is largely assembled rather than a vertically exploded view
- − The bread pieces on the sides look somewhat unnatural
FLUX.2 [dev] Turbo
- + Perfect adherence to the 'exploded' visual with vertically suspended layers
- + Accurate execution of the starburst for the price
- + Superior fiery, glowing effect on the text as requested
- − The meat patty texture is slightly less realistic compared to the other model
- − The sauce droplets look a bit painterly or stylized
Verdict: FLUX.2 [dev] Turbo followed the compositional instructions much better, providing a true vertically exploded view and including the requested starburst for the price. While FLUX.1 Kontext [max] has slightly sharper textures on the central burger, its failure to execute the specific layout and starburst element makes it the weaker choice for this precise prompt.
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 accuracy
- + Follows the menu item list perfectly including the completion of the 'Brown But...' cutoff
- + Crisp, high-contrast imagery
FLUX.2 [dev] Turbo
- + Superb realistic chalk texture with dusty smudges and varying pressure
- + Atmospheric cafe lighting and background Depth of Field
- + Dynamic handwriting style that feels more authentic to a real person
- − Has a slight logical error where it repeats the number '9' twice for the cookies
- − Text is slightly less legible than Model A's very clean script
Verdict: While FLUX.1 Kontext [max] delivered perfectly legible text and strictly followed the menu listing, FLUX.2 [dev] Turbo captured the requested 'chalk texture' and 'handwritten style' with much higher artistic realism. FLUX.2 [dev] Turbo is preferred because the text actually looks like it was written with physical chalk on a dusty board, despite a minor redundancy in the price rendering of the last item.
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 clean lighting and cinematic depth-of-field.
- + High anatomical accuracy for both the horse and the astronaut's suit.
- + Polished, professional digital art aesthetic with no visible artifacts.
- − Failed the negative prompt constraint: 'horse on top, not vice versa'.
- − Interpretation is fairly standard rather than truly surreal.
FLUX.2 [dev] Turbo
- + Ornate level of detail on the space suit and harness.
- + Dynamic background with diverse celestial bodies and atmospheric debris.
- + Rich textures in the horse's coat and mane.
- − Failed the negative prompt constraint: 'horse on top, not vice versa'.
- − Slightly less realistic anatomy in the horse's back legs.
Verdict: Both models failed to follow the logical inversion requested in the prompt ('horse on top'), instead providing a standard astronaut-riding-horse composition. FLUX.1 Kontext [max] is the preferred choice as it offers a much cleaner, more cinematic image with superior lighting and anatomical coherence compared to the slightly more cluttered FLUX.2 [dev] Turbo.
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 and lighting on the capybara.
- + Very cinematic bokeh for the Manhattan background.
- + Natural integration of the capybara into the car seat and seatbelt.
- − The passenger is out of focus and appears to be in the front seat or mid-car rather than the back.
- − Only one paw is clearly visible and it is not firmly on the steering wheel.
FLUX.2 [dev] Turbo
- + Perfectly adheres to the composition of having the passenger in the back seat.
- + Shows both front paws on the steering wheel as requested.
- + Captures the bored expression and phone usage of the businesswoman excellently.
- − The 'TAX' sign on the roof of the car is slightly misspelled/garbled.
- − The capybara's hands look a bit claw-like and less like natural paws.
Verdict: While FLUX.1 Kontext [max] has superior lighting and textures, it fails the spatial requirements of the prompt by placing the passenger in the front seat. FLUX.2 [dev] Turbo followed the prompt much more accurately, correctly placing the businesswoman in the back seat and showing both of the capybara's paws on the wheel.
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 typography style that feels premium and gothic.
- + Strong cinematic lighting with a very moody, dark atmosphere.
- + Beautifully detailed thorn and web border framing the piece.
- − Included redundant text in the location details at the bottom.
- − Used commas instead of periods in the date format requested.
- − Elements like the scroll ears on the main title are floating incorrectly.
FLUX.2 [dev] Turbo
- + Perfect adherence to all text fields including the date format and location.
- + Consistent and high-quality illustration of the bats and twisted trees.
- + Balanced composition with the scroll banner correctly placed under the pumpkin.
- − The 'parchment' texture is slightly less integrated into the background than the other model.
- − Lighting on the pumpkin feels a bit flatter compared to the atmospheric glow in the alternative.
Verdict: While FLUX.1 Kontext [max] has a more striking atmospheric glow and superior typography aesthetic, it fails on basic text accuracy by repeating the location and misformatting the date. FLUX.2 [dev] Turbo followed the prompt instructions perfectly, maintaining high visual quality and a more logical layout of elements like the scroll. FLUX.2 [dev] Turbo is the winner for its precision and clarity as a functional 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]
- + Excellent preservation of the jacket and background details.
- + Maintains the facial identity reasonably well despite slight compression.
- + The hair texture is detailed and thick as requested.
- − Significantly alters the lighting and shadows on the face, losing the original warmth.
- − The hairline integration with the forehead looks slightly artificial with a harsh edge.
FLUX.2 [dev] Turbo
- + Perfectly preserves the original lighting, skin tones, and background.
- + Highly realistic textured hair with soft, natural edges where it meets the face and background.
- + Creatively interprets 'full and thick' with a cohesive style that matches the beard texture.
- − The hair volume is very large, which might be more than some users intended for 'natural' density.
Verdict: FLUX.2 [dev] Turbo is the clear winner because it successfully added the hair while perfectly preserving the original image's lighting, color grade, and background. FLUX.1 Kontext [max] changed the overall exposure and lighting of the subject, making the edit look less like the original person and more like a new photograph altogether.
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 style with soft, rounded textures.
- + Clean, playful typography that matches the aesthetic.
- + High-quality rendering of the sauce and plate materials.
- − Missed the request for a small flag icon.
- − The isometric angle is slightly lower than the requested 45 degrees.
FLUX.2 [dev] Turbo
- + Includes all prompted elements including the flag icon.
- + Higher detail in the wood grain and rice textures.
- + Perfect 45-degree isometric composition and alignment.
- − The text is a bit plain and lacks the 'bold text' styling of the first model.
- − Visual style leans more realistic than 'cartoon scene'.
Verdict: FLUX.2 [dev] Turbo followed the prompt more comprehensively by including the flag icon and adhering strictly to the 45° isometric perspective. While FLUX.1 Kontext [max] had a more appealing 'cartoon' aesthetic, FLUX.2 [dev] Turbo's superior detail and prompt adherence make it the better overall result.
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]
- + Strongly preserves the subject's clothing style (denim shirt over black top).
- + Clear and simple caricature style with good linework.
- + Incorporates all elements: microphone/studio, dog, and hockey stick.
- − The facial likeness is significantly reduced, looking more like a generic cartoon character than the source person.
- − Exaggeration is a bit basic, mostly just larger eyes and glasses.
FLUX.2 [dev] Turbo
- + Excellent preservation of the subject's facial features and likeness within the caricature format.
- + Highly creative and detailed composition set in a hockey arena studio.
- + Stronger interpretation of the 'exaggerated and humorous' request with the oversized head and diverse dog characters.
- − Completely changes the subject's clothing from the source image to a formal suit.
- − Text on the desk is garbled and nonsensical.
Verdict: FLUX.2 [dev] Turbo is the superior choice for this challenge because it maintains a clear and recognizable likeness of the woman from the source image while following the complex prompt for a hockey-themed TV studio. While FLUX.1 Kontext [max] does a better job of preserving her specific denim outfit, its caricature style loses her identity, making her look like a generic character.
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 lighting requirements with prominent god rays and a warm golden glow.
- + Beautiful soft fur texture on all animals.
- + Charming, expressive eyes that look unified across the group.
- − The animals are static and sitting rather than 'tumbling' or 'playfully chasing' as requested.
- − The butterfly wings look slightly repetitive and less realistic in shape.
FLUX.2 [dev] Turbo
- + Captures the movement and 'tumbling' aspect of the prompt much better than Model A.
- + Highly detailed butterflies and visible dew sparkles that feel more tangible.
- + Better interaction between the animals, creating a more dynamic scene.
- − The kittens proportions and pose are slightly awkward as it leaps.
- − The lighting is a bit more muted compared to the 'masterpiece' glow of Model A.
Verdict: Both models followed the prompt closely, but FLUX.2 [dev] Turbo succeeded in capturing the 'tumbling' and active energy of the scene, whereas FLUX.1 Kontext created a more static, portrait-like group. FLUX.2 [dev] Turbo is preferred because it better translates the specific actions requested while maintaining high detail in elements like the dew and butterflies.
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 capture of modern Ghibli character design styles and line work
- + High source preservation including the exact shirt patterns and street layout
- + Strong color saturation consistent with the 'distracted boyfriend' meme aesthetic while being painterly
- − The faces feel slightly generic anime rather than specifically Ghibli
- − Less 'dreamy' background atmospheric effects compared to Model B
FLUX.2 [dev] Turbo
- + Perfect execution of the 'dreamy' and 'nostalgic' mood with cherry blossoms and soft lighting
- + Superior texture that mimics hand-painted watercolor on paper
- + Character faces have a distinct Ghibli-esque softness and rounded jawlines
- − The man's stubble feels slightly too detailed/realistic for the requested aesthetic
- − Deviates more from the urban setting of the source by adding heavy floral elements
Verdict: Both models successfully translated the meme into a Studio Ghibli style, but they chose different paths. FLUX.1 Kontext [max] prioritized keeping the exact urban environment and clothing patterns of the source image, resulting in a cleaner, modern anime look. FLUX.2 [dev] Turbo leaned much harder into the requested 'dreamy' and 'soft pastel' aesthetics by adding watercolor textures and cherry blossoms, making it the more visually evocative 'illustration' despite changing more of the background.
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 subject's face and original features
- + Natural and realistic wind effect integrated into the hair
- + Subtle and tasteful addition of flying leaves
- − Failed to preserve the pose of the left arm, moving it significantly from the original source
FLUX.2 [dev] Turbo
- + High adherence to the 'energetic' instruction with many flying leaves
- + Perfectly preserved the original pose and character layout
- + Dynamic, symmetrical hair motion
- − Some leaves appear to be floating awkwardly or overlapping the subject unnaturally
- − The hair rendering looks slightly more stylized and less natural than in the original source
Verdict: FLUX.1 Kontext [max] produced a more natural and photorealistic result, but it failed to preserve the original pose of the woman's arm. FLUX.2 [dev] Turbo followed the edit instructions more aggressively and maintained the source image's geometry perfectly, making it the better tool for precise instruction following despite some messy leaf artifacts.
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 period-accurate circumflex on 'CAFFÊ'
- + Clean woodcut-style illustration of the cloche
- + Perfect balanced composition following a vertical stack logic
- − The 'E' in 'Caffè' actually uses a circumflex rather than the grave accent requested by the prompt
FLUX.2 [dev] Turbo
- + Rich texture on the background mimics vintage paper well
- + Good use of space with additional decorative flourishes beneath the banner
- + Correct 'Caffè' grave accent
- − The 'EST.' banner is slightly distorted on the left side
- − The typography feels a bit thinner and less 'emblematic' than model a
Verdict: Both models followed the prompt exceptionally well, capturing the vintage aesthetic and required text. FLUX.1 Kontext [max] is the winner due to its cleaner vector-style lines and more robust, professional-looking typography, whereas FLUX.2 [dev] Turbo has a slightly warped banner and less impactful font choice.
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, large-scale layout.
- + Creative use of silhouettes for the crew members.
- + Stronger adherence to the 'flat-vector' aesthetic.
- − Missed several required infographic steps (Launch, Descent).
- − The rocket design is generic and does not resemble the Saturn V.
- − Some garbled text and illogical flow arrows.
FLUX.2 [dev] Turbo
- + Successfully included all requested steps including Launch, Earth Orbit, Translunar, Lunar Orbit, Descent, and Landing.
- + Better representation of the Saturn V and Lunar Module hardware.
- + Highly legible labels for each stage of the mission.
- − Layout is cluttered with repeated icons and redundant labels.
- − Took the prompt literally for 'icon' by including the word 'icon' in the text for the rocket.
- − One of the crew silhouettes appears to have a different hairstyle/gender (Collins).
Verdict: While FLUX.1 Kontext [max] has a much more professional and attractive aesthetic, it fails to follow the structured instructions for the infographic steps. FLUX.2 [dev] Turbo followed the prompt much more accurately, depicting every requested stage of the mission despite a more crowded and slightly less refined visual style. FLUX.2 [dev] Turbo is the winner for utility and prompt adherence.
Explore each model
Distilled version of Black Forest Labs' FLUX.2 [dev] outperforming it at a cheaper price. Developed by fal.ai.