Head to head
Esc

Models · slot A

to navigate to pick

FLUX.1 [dev] Black Forest Labs FLUX.2 [klein] 4B Black Forest Labs

Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.

FLUX.1 [dev]

24.6 arena score

#16 of 62 in Text-to-Image

Skill signature · Text-to-Image

FLUX.2 [klein] 4B

22.3 arena score

#32 of 62 in Text-to-Image

Vote tally

Where the votes landed

FLUX.1 [dev]

0%

win rate

Ties

0%

FLUX.2 [klein] 4B

0%

win rate

Shared challenges 13

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent handling of complex light refraction and reflections through the glass.
  • + The sphere appears to be levitating or suspended, creating a more artistic look.
  • + High-quality leather texture on the red book.
  • The glass object is constructed of thick glass slabs rather than a hollow cube, hindering the 'partially visible plant' prompt element.
  • The sphere is quite large, bordering on 'medium' rather than 'small' as requested.

FLUX.2 [klein] 4B

  • + Perfect adherence to the 'small' descriptor for the blue sphere.
  • + The glass cube is clearly hollow, allowing the plant and background to be seen through it as requested.
  • + Realistic grain texture on the wooden table.
  • The sphere's reflection on the bottom of the cube looks a bit more like a second sphere or a very sharp mirror rather than glass surface behavior.
  • The lighting on the book is slightly flat compared to the surrounding environment.

Verdict: FLUX.1 [dev] produces a more aesthetically pleasing and high-end photographic result with superior textures, but it struggles with the physics of the 'hollow cube' requested. FLUX.2 [klein] 4B follows the spatial instructions more accurately, correctly placing a small sphere inside a hollow glass container and showing the plant through the glass.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent composition with a strong use of 50mm shallow depth of field.
  • + Very high visual quality with realistic skin textures and rain effects.
  • + The character's interaction with the bicycle feels deliberate and grounded.
  • Lacks the requested 'motion blur' from passing cars.
  • The environment looks slightly generic rather than specifically Japanese.

FLUX.2 [klein] 4B

  • + Captures the 'imperfect framing' and 'candid' street vibe more effectively.
  • + Better adherence to the 'motion blur' request in the background vehicles.
  • + Authentic depiction of a traditional Japanese utility bicycle.
  • Anatomical issues where the man's leg and bicycle frame intersect awkwardly.
  • Physics of the bicycle chain and kickstand area are inconsistent.

Verdict: FLUX.1 [dev] produces a cleaner, higher-quality image with superior lighting and depth, though it misses specific technical prompts like motion blur. FLUX.2 [klein] 4B follows the stylistic prompt more closely with a authentic candid feel, but suffers from significant geometric clipping and structural errors in the bicycle and character's pose. FLUX.1 [dev] is the preferred choice for its overall coherence and photographic realism.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent shallow depth of field and bokeh quality.
  • + Highly detailed, lifelike iris texture in the eyes.
  • + Captures natural skin texture including subtle freckles and pores.
  • Missed the request for small beads in the hair braids.
  • Subject looks more like a modern woman in cosplay than a 'battle-worn paladin'.
  • Armor engraving is fairly plain compared to the prompt's request.

FLUX.2 [klein] 4B

  • + Perfect adherence to the hair beads, scars, and ornate engraving requirements.
  • + Excellent lighting setup with visible torches and warm reflections.
  • + Captures the 'battle-worn' aesthetic much more convincingly.
  • The 'small beads' in the hair slightly warp the hair geometry around them.
  • Skin texture is a bit flat compared to the realism of Model A.

Verdict: FLUX.2 [klein] 4B is the clear winner as it followed every detail of the prompt, including the specific requests for hair beads and ornate engravings which FLUX.1 [dev] ignored. While FLUX.1 [dev] has slightly more realistic skin and eyes, it failed to capture the 'battle-worn' character essence, looking too clean and simplified.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent typographic clarity and hierarchy
  • + Highly professional minimalist aesthetic suitable for a real restaurant
  • + Very clean layout with effective use of white space
  • Technically failed the 'grid of food photos' prompt by using only two circular images
  • Did not include a specific 'Pizza' section heading

FLUX.2 [klein] 4B

  • + Successfully implemented the 'grid of food photos' request
  • + Vibrant food photography and colorful accents
  • + Captures the high-energy casual dining feel well
  • Text rendering is poor with many gibberish characters and overlapping letters
  • The layout feels cluttered compared to Model A
  • Font choices are less legible and consistent

Verdict: FLUX.1 [dev] produces a much more professional and realistic menu layout with clean typography, though it missed the specific instruction for a grid of photos. FLUX.2 [klein] 4B adhered better to the 'grid' and 'vibrant' aspects of the prompt, but the actual text rendering and overall design feel chaotic and difficult to read. FLUX.1 [dev] is the superior choice for a professional design task despite the minor prompt omission.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent food photography style with realistic textures
  • + Accurate limited time and price text placement
  • + Clean, balanced composition capturing the 'exploded' view
  • Failed to include the main title 'MAGIC BURGER' entirely
  • Missing the requested starburst for the price

FLUX.2 [klein] 4B

  • + Stronger sense of motion and 'fire' theme with dynamic embers
  • + Includes the starburst element for the price as requested
  • + Rendered the 'Limited Time Only' text clearly
  • The main title text is severely mangled and misspelled
  • Failed the 'exploded' instruction as the burger is mostly assembled
  • The bun texture looks plastic and less photorealistic than its competitor

Verdict: FLUX.1 [dev] produced a much higher quality image in terms of food aesthetics and photorealism, though it failed to include the primary brand name 'Magic Burger'. FLUX.2 [klein] 4B attempted more elements of the prompt like the starburst and embers, but the burger is not properly exploded and the main text is unintelligible. FLUX.1 [dev] is the winner for professional visual quality despite the missing text.

Chalkboard Menu

Text-to-Image

“Handwritten-style chalkboard menu in a cozy café, all text rendered in the exact same realistic chalk handwriting style with natural variations in letter size, slight slant, and chalk texture — no printed or digital fonts anywhere on the board. Title at the top in elegant cursive chalk handwriting: ‘TODAY’S SPECIALS – APRIL 30, 2026’. Below it, three menu items also in the same handwritten chalk style: ‘Truffle Mushroom Risotto – $24’, ‘Grilled Octopus with Lemon & Herbs – $28’, ‘Brown Butter Chocolate Chip Cookies – $9’. At the very bottom, smaller text in the identical handwritten chalk style (slightly smaller but still clearly legible with the same handwriting characteristics): ‘All items made fresh daily • Ask about our gluten-free options’. Warm ambient café lighting, visible chalk dust and smudges, realistic handwriting imperfections, no clean printed text anywhere.”

FLUX.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Perfect spelling on all requested menu items.
  • + Consistent and legible handwriting style.
  • + Strong background depth with a realistic café environment.
  • The text looks a bit more like a digital brush than raw chalk.
  • The handwriting is almost too perfect, lacking significant natural variation.

FLUX.2 [klein] 4B

  • + Excellent chalk texture including dust and smudges.
  • + Realistic handwriting variation and slant.
  • + Captures the cursive requirement for the title well.
  • Multiple spelling errors throughout the text (e.g., 'Truffel', 'Musheram', 'Ootrpous').
  • Spacing issues in the main title ('S PECIALS').

Verdict: FLUX.1 [dev] followed the text instructions perfectly, producing an error-free and highly legible menu, though the chalk texture is slightly clean. FLUX.2 [klein] 4B achieved a much better artistic 'chalk' feel with realistic smudging, but it failed significantly on spelling and character spacing. FLUX.1 [dev] is the clear winner for its superior prompt adherence regarding the specific text requested.

The Reversed Rodeo

Text-to-Image

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

FLUX.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Successfully followed the difficult constraint of 'horse on top' of the astronaut.
  • + Maintains a clean, cinematic aesthetic with smooth lighting.
  • + Shows high creativity in interpreting a surreal physical arrangement.
  • The horse's anatomy is slightly simplified and rubbery.
  • Lighting on the horse's mane is a bit inconsistent with the background light source.

FLUX.2 [klein] 4B

  • + Excellent texture rendering on the spacesuit and horse's coat.
  • + Strong composition with realistic atmospheric perspective over the planet.
  • Failed the primary prompt constraint by placing the astronaut on top of the horse.
  • Missing the requested 'surreal' aspect, opting for a standard sci-fi interpretation.
  • The horse's front legs exhibit anatomical clipping issues.

Verdict: FLUX.1 [dev] is the winner because it successfully followed the specific prompt instruction to place the horse on top of the astronaut, creating a truly surreal image. FLUX.2 [klein] 4B ignored this spatial instruction entirely, producing a standard astronaut riding a horse, and also suffered from some anatomical distortion in the horse's legs.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent photorealistic texture on the capybara's fur.
  • + Great depth of field with cinematic lighting and colors.
  • + Follows the 'yellow taxi driver cap' prompt perfectly.
  • The woman appears to be in the front passenger seat rather than the back seat.
  • The paws on the steering wheel look slightly AI-generated and less defined.
  • The composition feels more like a passenger-side view than a full taxi interior.

FLUX.2 [klein] 4B

  • + Correctly places the passenger in the back seat as requested.
  • + The capybara's paws are very clearly placed on the steering wheel.
  • + Captures the 'bored' expression of the businesswoman more effectively.
  • The car interior looks slightly cramped and less like a modern NYC taxi.
  • The lighting is a bit flat compared to the cinematic quality of the first image.
  • The capybara's jacket is a bit bulky and interferes with the steering wheel visual.

Verdict: FLUX.1 [dev] produced a more visually stunning and photorealistic image with beautiful lighting, but it failed to follow the spatial instruction of placing the woman in the back seat. FLUX.2 [klein] 4B followed the prompt's compositional requirements much more accurately, showing a clear separation between the driver and the backseat passenger, making it the better conceptual match despite slightly lower aesthetic polish.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Strong composition with a clear focal jack-o-lantern
  • + Clean rendering of the thorn border
  • + Excellent contrast and cinematic lighting
  • Major spelling errors in the main title
  • Redundant text listing 'Time' twice and adding commas in the date
  • Lack of visible cobwebs asked for in the prompt

FLUX.2 [klein] 4B

  • + Successfully included all elements including cobwebs and the moody night sky
  • + More accurate 'vintage gothic' parchment aesthetic
  • + The banner text is almost perfectly spelled
  • The main title text is severely garbled and illegible
  • The date has a typo ('206' instead of '2026')
  • Visuals are slightly more cluttered compared to model A

Verdict: Both models struggled significantly with the specific text requirements, but FLUX.1 [dev] produced a more polished and professional-looking layout with superior lighting, despite the 'Falloween Rantcy' typo. FLUX.2 [klein] 4B better captured the requested gothic atmosphere with webs and parchment textures, but the large distorted title text makes it less usable for a real invitation.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent diorama-style composition with a tiered base
  • + High-quality soft lighting and 3D materials
  • + Includes a clean Japan flag icon and creative garnish
  • Text contains a hallucination ('SUSH CATON')
  • Depth of field is a bit shallow, blurring the back sushi pieces

FLUX.2 [klein] 4B

  • + Text 'JAPAN' is very large and bold as requested
  • + Render quality on the sushi fish is highly realistic with PBR-like properties
  • + Minimalistic and clean layout
  • Failed to include the 'I' in 'SUSHI', rendering 'SUSH'
  • Flag icon is incorrect (appears as the flag of Austria)
  • Missed the 'raised diorama base' requirement, using just a plate

Verdict: FLUX.1 [dev] followed the 'isometric miniature diorama' aesthetic much better, creating a cohesive 3D scene. While it had a minor spelling error in the subtext, FLUX.2 [klein] 4B failed more significantly by misspelling the primary word 'SUSHI' and generating an incorrect flag.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Strong bokeh and warm atmospheric lighting
  • + High visual appeal in a stylized, 3D animated movie aesthetic
  • Failed to include a kitten and a bunny, generating four generic canine-like creatures instead
  • Completely ignored the hyper-photorealistic requirement in favor of a cartoon/Pixar style
  • Anatomically incorrect paws that look like mittens

FLUX.2 [klein] 4B

  • + Excellent adherence to the 'hyper-photorealistic' request with realistic fur textures
  • + Includes all requested animals (puppy, kitten, fox) correctly
  • + Beautiful implementation of god rays and dew sparkles as requested in the prompt
  • The bunny is missing from the group
  • One kitten appears much smaller and slightly warped in its positioning

Verdict: FLUX.2 [klein] 4B is the clear winner for its superior realism and adherence to the specific animal types mentioned in the prompt. While FLUX.1 [dev] produced a cute image, it completely ignored the 'photorealistic' style instruction and failed to differentiate between the species of animals, resulting in four very similar-looking cartoonish creatures.

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.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent typography style that fits the vintage aesthetic
  • + Clean cloche icon with elegant steam lines
  • + Includes a well-designed banner element
  • Significant spelling errors ('Flariláan' instead of 'Florian', 'Reseaurant')
  • Includes random, non-requested numbers (11011, 1941)

FLUX.2 [klein] 4B

  • + Beautiful parchment-style texture on the background
  • + Cloche dome has nice lighting and dimension
  • + Consistent color palette throughout
  • Spelling error in the main brand name ('Flaxtion' instead of 'Florian')
  • Redundant repetition of 'Est. 1720' both inside and below the banner
  • Font choice is a bit generic for a luxury 'Caffè' logo

Verdict: Both models failed to spell the primary brand name correctly, but FLUX.1 [dev] produced a more cohesive and professional-looking vintage logo design. While FLUX.2 [klein] 4B followed the texture and layout instructions well, its redundancy in text and less sophisticated typography make it the weaker 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.”

FLUX.1 [dev]
FLUX.2 [klein] 4B

AI Judge Analysis

FLUX.1 [dev]

  • + Excellent adherence to the muted NASA-inspired color palette.
  • + Sophisticated layout that feels like a professional infographic poster.
  • + Clean vector lines and consistent iconography across the design.
  • Text rendering is mostly gibberish despite the header being correct.
  • Includes irrelevant planets like Saturn which were not part of the mission request.

FLUX.2 [klein] 4B

  • + Clearly labels distinct steps of the mission in a sequential grid.
  • + Good rendering of the Lunar Module and the Tranquility base pin.
  • + Follows the specified 6-step structure more literally than model A.
  • Severe spelling errors in the main title ('APLLO 1UT 1J').
  • The Saturn V rocket illustration is anatomically incorrect and messy around the engines.
  • Visual style is less cohesive, with varying icon sizes and awkward spacing.

Verdict: FLUX.1 [dev] produces a significantly more professional-looking infographic with superior composition and color usage, though it struggles with logical mission steps and text. FLUX.2 [klein] 4B follows the prompt's structural instructions more closely but fails significantly on basic typography and aesthetic polish. FLUX.1 [dev] is the preferred choice for its high-quality vector art style that better matches the 'modern infographic' requirement.

Next steps

Explore each model