Head to head
Esc

Models · slot A

to navigate to pick

FLUX.2 [dev] Turbo fal FLUX.2 [max] Black Forest Labs

Settled by community votes across 20 shared challenges, with an AI judge weighing in on each.

FLUX.2 [dev] Turbo

27.8 arena score

#4 of 62 in Text-to-Image

Skill signature · Text-to-Image

FLUX.2 [max]

25.9 arena score

#10 of 62 in Text-to-Image

Vote tally

Where the votes landed

FLUX.2 [dev] Turbo

66.7%

win rate

Ties

0.0%

FLUX.2 [max]

33.3%

win rate

66.7% 0.0% ties 33.3%
Shared challenges 20

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]
67% wins 0% ties 33% wins

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent realism in textures, particularly the grain of the wooden table and the dust on the glass.
  • + Perfect adherence to the lighting instruction with a soft, natural window light from the left.
  • + Highly believable depth and refraction of the plant through the glass cube.
  • The glass cube looks more like a small aquarium or hand-joined glass rather than a sleek minimalist cube.

FLUX.2 [max]

  • + Clean, modern aesthetic with sharp geometric lines for the cube.
  • + Strong color saturation on the red book and blue sphere.
  • + Includes a reflective base inside the cube that adds visual interest.
  • Lighting is somewhat flat and lacks the 'soft window light' realism compared to the other model.
  • The background plant is very blurry, making it harder to see through the glass as requested.

Verdict: FLUX.2 [dev] Turbo produced a significantly more realistic image with convincing textures and perfect lighting that matches the prompt's request for soft window light. While FLUX.2 [max] has a cleaner, more graphic design look, it feels more like a 3D render and lacks the natural photographic quality of the Turbo model.

Man and Car in California

Editing
Edit instruction

“Make a photo of the man driving the car down the California coastline”

Source
FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent character preservation of the man's face and specific clothing (plaid coat and scarf).
  • + Immersive composition with motion blur and a realistic interior perspective.
  • + High visual quality with realistic lighting on the car and figure.
  • The car color has shifted from the original bright white to a more metallic silver/beige.

FLUX.2 [max]

  • + Perfectly preserves the original white color and model of the Rolls-Royce.
  • + Accurately recreates the California coastline environment and highway markings.
  • + Preserves the man's hair and clothing style effectively.
  • The scale of the man inside the car is slightly off, making him look a bit small for the seat.
  • The face details are less consistent with the source image than in Model A.

Verdict: Both models performed exceptionally well at combining the two source images. FLUX.2 [dev] Turbo created a more dynamic and cinematically pleasing shot with better facial likeness, though it altered the car's color. FLUX.2 [max] was more faithful to the original car's color and external characteristics, making it the better choice for strict object preservation.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the 'imperfect framing' prompt with a slightly high and wide angle.
  • + Very realistic skin textures and hand details for an elderly subject.
  • + Captures the motion blur of passing cars effectively while maintaining focus on the subject.
  • The front wheel of the bicycle is clipping into the pavement.
  • The bicycle's anatomy overall is slightly more jumbled and cluttered compared to Model B.

FLUX.2 [max]

  • + Exceptional realism in the textures of the jacket and the wet pavement.
  • + The mechanical details of the red bicycle are very coherent and well-constructed.
  • + Beautiful bokeh and shallow depth of field that perfectly matches the 50mm lens request.
  • The composition feels a bit more 'staged' and professional, missing the 'imperfect framing' requested.
  • The subject's hands are slightly merged with the spokes of the wheel.

Verdict: Both models handled the complex request for a rainy street scene with high fidelity. FLUX.2 [dev] Turbo followed the 'imperfect framing' and 'candid' aspects of the prompt more closely, resulting in a shot that looks like an authentic street photo, despite a clipping error with the front wheel. FLUX.2 [max] produced a more polished, cinematic image with superior technical rendering of materials and light, though it feels slightly more posed.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the 'beads in hair' prompt with colorful, distinct elements.
  • + Superior texture on the leather straps and buckles, showing realistic wear and grain.
  • + Powerful lighting composition with the torch visible and its light wrapping naturally around the subject.
  • The sparks appear a bit more like static noise or digital artifacts than embers.
  • The facial scars look a bit like surface paint rather than physical indentations in the skin.

FLUX.2 [max]

  • + Very high-quality facial features with a more intense, lifelike expression and realistic scarring.
  • + Elegant engraving details on the plate armor that look deeply etched and high-quality.
  • + Beautiful bokeh effect with softer, more natural-looking light particles.
  • The beads in the hair are much more subtle and less distinct than requested.
  • The armor engraving is slightly repetitive compared to the more varied patterns in Model A.

Verdict: Both models performed exceptionally well, but FLUX.2 [dev] Turbo followed the specific textural prompts for leather and hair beads more accurately. FLUX.2 [max] achieved a more cinematic and intense facial expression with superior skin rendering, but FLUX.2 [dev] Turbo is the winner for its better lighting composition and comprehensive adherence to all detailed prompt elements.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Uses high-quality, realistic food photography that feels professional.
  • + Strong, clean header with clear sans-serif typography.
  • + Excellent use of vibrant secondary colors in the border and accents.
  • Logical inconsistency in pricing (e.g., $260 for a pizza).
  • Text layout becomes messy in the 'Mains' section with overlapping characters.

FLUX.2 [max]

  • + Perfectly organized grid layout that balances photos and text effectively.
  • + More realistic pricing structure for a casual dining establishment.
  • + Highly consistent minimalist design with color-coordinated section headers.
  • Some food images are slightly repetitive or lack the high-end polish of Model A.
  • Small decorative icons are a bit generic.

Verdict: Both models followed the prompt exceptionally well, but FLUX.2 [max] produced a more functional and logically coherent menu layout with realistic pricing and perfectly aligned sections. While FLUX.2 [dev] Turbo had superior individual food photography and a bold header, its internal text alignment and nonsensical prices ($260 pizza) made it less successful as a design prototype.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent text integration with consistent fiery glow effects across all three required text elements.
  • + Highly detailed and realistic burger textures, particularly the patty and the dripping sauce.
  • + Strong sense of 'exploded' depth with ingredients clearly suspended in a 3D space.
  • The bun looks slightly dry/matte compared to the vibrant toppings.

FLUX.2 [max]

  • + Beautiful lighting on the top bun giving it a glossy, appetizing look.
  • + Dynamic use of particles and crumbs to enhance the sense of motion.
  • + Excellent clarity on the '€6.99' starburst text.
  • The 'exploded' effect is less pronounced, as the top half of the burger remains largely assembled.
  • The starburst element lacks the 'fiery, glowing effect' requested, looking more like a flat graphic.

Verdict: Both models followed the prompt well, but FLUX.2 [dev] Turbo produced a better overall advertisement by applying the requested fiery glow effect to all text elements, including the starburst. FLUX.2 [max] had better lighting on the bread, but failed to separate the burger components as effectively as the Turbo model, resulting in a less 'exploded' look.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent chalk texture with realistic smudges and dusty residue
  • + Accurate rendering of the requested text and prices
  • + Very realistic background depth-of-field and café atmosphere
  • The layout for the octopus item is a bit awkward with the hyphen placement
  • Lettering style is slightly inconsistent between the title and the body

FLUX.2 [max]

  • + Perfectly polished and centered layout with very legible handwriting
  • + Includes the full 'Brown Butter' item requested in the prompt
  • + Superior lighting and framing with a clean wooden border
  • Texture looks a bit too clean/digital for raw chalk
  • Handwriting is very uniform, losing some of the 'natural variation' requested

Verdict: Both models handled the complex text requirements exceptionally well. FLUX.2 [dev] Turbo offers a more authentic, gritty chalk texture that feels hand-drawn, whereas FLUX.2 [max] provides a cleaner, more aesthetically balanced composition with perfect formatting. FLUX.2 [max] is the slight winner for completing the 'Brown Butter' item more logically within the frame.

Pose & Character Mashup

Editing
Edit instruction

“Use Image 1 as the exact pose reference and Image 2 as the character reference. Recreate the person/character from Image 2 in the exact dynamic pose and body position from Image 1. Keep the exact face, hair, clothing style/details, and expression from Image 2. Match the lighting and environment of Image 1. The final image must show the character from Image 2 performing the precise action/pose from Image 1 with perfect anatomy and natural integration.”

Source
FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the complex posing of the head and body from the source image.
  • + Accurately recreates the character's clothing details including the specific scarf pattern and shirt text.
  • + Effectively captures the lighting and color grade of the original environment.
  • Gave the character long feminine hair that was not present in the character reference.
  • Failed to include the character's sunglasses.
  • The foot on the box has anatomical issues with toe placement.

FLUX.2 [max]

  • + Perfectly captures the character's facial features, hair, and sunglasses from Image 2.
  • + Maintains the red ottoman and yellow background from the source environment.
  • + Clothing details like the scarf and shirt text are highly accurate to the reference.
  • Completely failed to replicate the specific dynamic pose from Image 1, choosing a generic squat instead.
  • The anatomy of the feet and legs is distorted and unclear.
  • The outstretched arm position significantly deviates from the reference.

Verdict: FLUX.2 [dev] Turbo followed the posing instructions much better, capturing the difficult lean and head angle of Image 1, though it incorrectly changed the character's hair length and removed his sunglasses. FLUX.2 [max] achieved a near-perfect character likeness but failed the primary task of recreating the specific 'exact dynamic pose' from Image 1, resulting in a different body position entirely. FLUX.2 [dev] Turbo is preferred for its superior spatial and structural adherence to the source image.

The Reversed Rodeo

Text-to-Image

“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent sense of motion and dynamic composition
  • + Highly cinematic lighting with a multi-layered planet background
  • + Detailed textures on the spacesuit and horse's coat
  • Prompt adherence failure: shows astronaut on top of horse
  • Anatomical issues with the horse's back right leg

FLUX.2 [max]

  • + Sharp rendering and high resolution overall
  • + Creative use of a floating asteroid as a pedestal
  • + Good lens flare effects and starry background
  • Prompt adherence failure: shows astronaut on top of horse
  • The astronaut's leg position is awkward and merges slightly with the saddle

Verdict: Both models failed the specific spatial logic of the prompt, which requested the horse to be on top of the astronaut ('horse on top, not vice versa'). Since both produced the standard 'astronaut on horse' trope, FLUX.2 [dev] Turbo is the winner due to its superior cinematic composition, sense of scale with the planet, and more dynamic movement compared to the static pose in FLUX.2 [max].

Outfit Transfer Challenge

Editing
Edit instruction

“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”

Source
FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Successfully transferred almost all clothing and accessories from Image 2.
  • + High level of detail in the added jewelry and textures.
  • Failed completely to keep the base person from Image 1, replacing him with the man from Image 2.
  • Changed the person's face, hair, and skin characteristics.

FLUX.2 [max]

  • + Perfectly preserved the identity, face, hair, and unique skin patterns of the person in Image 1.
  • + Accurately adapted the clothing from Image 2 to the body and pose of the person in Image 1.
  • + Maintained the original background and lighting conditions.
  • Missed some of the more elaborate jewelry seen in the scarf area of Image 2.
  • The watch and ring are present but slightly different from the source.

Verdict: FLUX.2 [max] followed the complex editing instructions nearly perfectly, maintaining the identity of the person in Image 1 while dressing them in the requested outfit. FLUX.2 [dev] Turbo failed the primary instruction by simply replacing the person from Image 1 with the person from Image 2.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent character expressions that match the 'bored' and 'professional' descriptions perfectly
  • + High-quality fur texture and realistic lighting on the subjects
  • + Better adherence to the 'inside a taxi' perspective by showing the dash and windshield
  • The passenger is sitting in the front passenger seat rather than the back seat as requested
  • The taxi sign is visible through the roof in an anatomically impossible way for a car interior

FLUX.2 [max]

  • + Correctly places the passenger in the back seat
  • + Fantastic detail in the car interior including dashboard lights and tactile materials
  • + Highly realistic profile of the capybara in clothing
  • The passenger's expression is slightly more focused/scowling than bored
  • The side-view composition feels a bit cramped compared to the wide view of the city

Verdict: Both models followed the complex prompt with high fidelity, but FLUX.2 [max] succeeded in placing the passenger in the back seat, whereas FLUX.2 [dev] Turbo placed her in the front. FLUX.2 [dev] Turbo had slightly better 'bored' expression work for the businesswoman, but FLUX.2 [max] is the superior image due to correct spatial reasoning and a more detailed car interior.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the 'dark parchment' texture and border request.
  • + Perfect text rendering with consistent gothic font styles.
  • + Very detailed vine and web border that frames the central pumpkin perfectly.
  • The bats look a bit stiff and symmetrical.
  • The pumpkin is slightly smaller compared to the overall layout.

FLUX.2 [max]

  • + Dynamic lighting on the pumpkin is very effective.
  • + The gothic title has a nice glow effect.
  • + Better sense of atmospheric depth in the background.
  • The parchment texture is less prominent than requested.
  • The 'You are invited' scroll banner has slightly inconsistent perspective compared to the foreground.
  • The spider web corner elements feel more transparent/faded than Model A.

Verdict: Both models followed the prompt exceptionally well, producing high-quality text and thematic imagery. FLUX.2 [dev] Turbo is the winner because it better captures the 'dark parchment' and 'spooky border' requirements, creating a more cohesive vintage invitation feel, whereas FLUX.2 [max] feels a bit more like a modern digital painting.

Bald man challenge

Image Editing
Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After
FLUX.2 [dev] Turbo
Before After
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent preservation of the original facial features and lighting.
  • + Naturally blends the new hair texture into the existing sideburns and beard.
  • + Maintains very high resolution and detail in the added hair fibers.
  • The Choice of a large afro-style texture may be seen as a less 'natural' transition for the subject's existing hair type.
  • The hair volume is slightly disproportionate to the head size.

FLUX.2 [max]

  • + Creates a very realistic and fitting hairstyle for the man's existing beard and features.
  • + Flawless integration of the hairline with the forehead and eyeglasses.
  • + Perfectly preserves the background, clothing, and lighting of the source image.
  • Minor smoothing of the forehead skin texture compared to the source image.

Verdict: Both models did an exceptional job of preserving the source image while adding hair. FLUX.2 [dev] Turbo added a wilder, curly texture that looks slightly like a wig due to its massive volume, whereas FLUX.2 [max] provided a more believable, stylish, and naturally integrated head of hair that feels like a true extension of the original person.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent high-fidelity textures on the salmon and rice grains.
  • + Perfectly renders the text and flag icon with high clarity.
  • + Realistic lighting and shadows create a strong 3D effect.
  • The 'miniature 3D cartoon' style leans more towards realism than a stylized diorama.
  • The perspective on the 'JAPAN' text feels slightly detached from the 3D scene.

FLUX.2 [max]

  • + Stronger adherence to the 'miniature 3D cartoon scene' and 'diorama' aesthetic.
  • + Includes more variety in the sushi types, including rolls and nigiri.
  • + Very clean, soft PBR materials that match the requested cartoon style.
  • The chopsticks are fused together and lack detail.
  • Some of the sushi textures look a bit flat compared to the realism request.

Verdict: FLUX.2 [dev] Turbo produced an image with stunning realism and material detail, while FLUX.2 [max] better captured the 'miniature diorama' and 'cartoon' aspect of the prompt. FLUX.2 [dev] Turbo is the preferred choice for its superior text rendering and impressive food textures that make the scene pop.

Over-the-top cartoon caricature

Editing
Edit instruction

“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.”

Source
FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent preservation of the subject's facial features and eye color from the source image.
  • + Includes detailed hockey background elements like players and multiple screens.
  • + The illustration style has a high-quality, professional caricature feel with realistic shading.
  • The transition from the tiny neck to the suit is a bit awkward.
  • The hands are slightly stylized to the point of looking a bit flat.

FLUX.2 [max]

  • + Successfully incorporates all elements: tv anchor, dogs, and hockey in a cohesive Disney-like style.
  • + Preserves the denim shirt from the original source image.
  • + Very clean and vibrant vector-style illustration.
  • The facial resemblance to the source image is weak, changing the eye color and facial structure significantly.
  • The caricature is more of a generic cartoon character than an exaggerated version of the specific woman provided.

Verdict: FLUX.2 [dev] Turbo is the winner because it successfully creates a caricature that actually looks like the woman in the source image, capturing her specific smile and brown eyes. While FLUX.2 [max] produces a very clean and fun illustration, it loses the subject's likeness, creating a generic blue-eyed character that does not feel like an edit of the original person.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent depiction of dew sparkles and golden hour lighting.
  • + Highly expressive and joyful facial expressions on the animals.
  • + Rich, detailed textures on the fur of each animal.
  • The kitten has an anatomically strange extra-long tail.
  • The interaction feels slightly cluttered with animals overlapping tightly.

FLUX.2 [max]

  • + Superior composition with a clear sense of movement and 'chasing'.
  • + More accurate lighting with distinct god rays and atmospheric depth.
  • + Consistent anatomy across all four animals.
  • The animals' expressions are slightly more vacant compared to Model A.
  • The bunny's texture is a bit smoother and less detailed than the other animals.

Verdict: Both models followed the complex prompt perfectly, including all four specific animals. FLUX.2 [dev] Turbo produced more emotive faces and beautiful dew effects, but FLUX.2 [max] succeeded with a much better composition that actually showed the animals running and chasing butterflies, whereas the animals in the first image are mostly sitting or tumbling in place. FLUX.2 [max] is the winner for its better anatomy and more dynamic scene arrangement.

Studio Ghibli Anime Style

Editing
Edit instruction

“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”

Source
FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to Ghibli style with clean linework and watercolor-inspired shading.
  • + Preserves the original composition and character poses nearly perfectly.
  • + Beautifully reimagined background with cherry blossoms and soft, dreamy lighting.
  • The man's facial expression is slightly more neutral/sad than the original's whistling/pouting look.

FLUX.2 [max]

  • + Successfully captures a warm, nostalgic mood with soft textures.
  • + Preserves the source image's layout and the specific 'distracted' expression of the man well.
  • + Good use of pastel tones and painterly washes.
  • The character designs feel a bit less 'Ghibli-esque' and more generic anime compared to Model A.
  • Fewer interesting background details added compared to the cherry blossoms in the other version.

Verdict: Both models did an exceptional job at translating a well-known meme into a specific art style while maintaining the original's composition. FLUX.2 [dev] Turbo (Model A) is the winner because its interpretation of the Studio Ghibli aesthetic is more authentic, featuring beautiful floral background details and superior linework that feels more like hand-painted cel art, whereas FLUX.2 [max] (Model B) feels slightly more faded and less detailed.

Golden Hour Stroll

Image Editing
Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

Before After
FLUX.2 [dev] Turbo
Before After
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Successfully added a large number of falling leaves across the entire composition.
  • + Accurately modified the hair to flow outward in multiple directions.
  • The hair edit looks slightly symmetrical and artificial compared to the source.
  • A leaf is rendered directly on top of the woman's face/chin.

FLUX.2 [max]

  • + The hair motion looks more natural and directional, as if blown by a specific gust of wind.
  • + Excellent preservation of the woman's face and the dog's features.
  • + The leaves have varied sizes and focus, creating a better sense of depth.
  • One cluster of leaves on the left has a slightly repetitive or cloned appearance.

Verdict: Both models followed the instructions well, but FLUX.2 [max] produced a more believable and aesthetically pleasing result. While FLUX.2 [dev] Turbo added more leaves, it placed one awkwardly over the subject's face, whereas FLUX.2 [max] managed hair physics more realistically and maintained better clarity on the subjects.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent typography with strong visual hierarchy and perfect spelling.
  • + The banner and cloche illustrative style perfectly capture the 'vintage minimalist' request.
  • + Superior use of texture and warm tones creating a high-quality paper feel.
  • The steam lines are a bit thick compared to the delicate nature of the rest of the logo.

FLUX.2 [max]

  • + Clean circular emblem layout that feels well-balanced.
  • + Accurate text placement and correct spelling.
  • + Good adherence to the requested color palette.
  • The illustrative style of the cloche is a bit generic and clinical.
  • The banner ribbon is quite small and lacks the 'vintage' detail seen in the other version.

Verdict: Both models followed the prompt perfectly, including difficult details like the accent on 'Caffè'. FLUX.2 [dev] Turbo is the winner because its composition and typography feel like a professional boutique logo, whereas FLUX.2 [max] feels slightly more like a generic clip-art emblem.

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.”

FLUX.2 [dev] Turbo
FLUX.2 [max]

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + High level of technical detail in the illustrations of the Saturn V and Lunar Module.
  • + Excellent inclusion of crew portraits with accurate names and silhouettes.
  • + Comprehensive layout that feels like a detailed educational poster.
  • The layout is cluttered and the flowchart logic is confusing and redundant.
  • Included literal prompt text like 'Saturn Vicon' instead of just the icon.
  • Failed to maintain a 'flat-vector' style, opting for more realistic textures and 3D-looking elements.

FLUX.2 [max]

  • + Perfect adherence to the 'flat-vector' style and 'clean, modern' aesthetic.
  • + Logical, easy-to-follow grid layout for the mission steps.
  • + Accurate iconography that maintains consistent line weights and styling across the set.
  • Minor spelling error in the word 'Tranquiity'.
  • Missing one of the requested steps (it only shows 5 icons for 6 steps).
  • Translunar icon is slightly abstract compared to the others.

Verdict: FLUX.2 [max] followed the stylistic instructions much better, producing a clean, flat-vector infographic that looks professionally designed despite a minor typo. FLUX.2 [dev] Turbo provided more detailed technical illustrations but failed the 'flat' style requirement and produced a messy, repetitive layout with text artifacts from the prompt.

Next steps

Explore each model