Head to head
Esc

Models · slot A

to navigate to pick

FLUX.2 [dev] Black Forest Labs Recraft V4 Recraft AI

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

FLUX.2 [dev]

24.3 arena score

#19 of 62 in Text-to-Image

Skill signature · Text-to-Image

Recraft V4

25.6 arena score

#11 of 62 in Text-to-Image

Vote tally

Where the votes landed

FLUX.2 [dev]

33.3%

win rate

Ties

0.0%

Recraft V4

66.7%

win rate

33.3% 0.0% ties 66.7%
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.2 [dev]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Perfect adherence to spatial instructions with the sphere resting on the bottom of the cube.
  • + Excellent realism in the texture of the red book and the wooden table grain.
  • + Very natural lighting and reflections that match the 'left window' source description.
  • The glass cube has quite thick edges which slightly obscures the sphere.

Recraft V4

  • + Exceptional glass rendering with high clarity and convincing refraction.
  • + Beautiful bokeh effect on the plant in the background.
  • + High level of detail on the paper texture of the book pages.
  • The blue sphere is levitating inside the cube rather than resting on the surface.
  • The scale of the sphere is slightly larger than 'small' relative to the cube.

Verdict: FLUX.2 [dev] followed the prompt's logical spatial arrangement better, placing the sphere realistically inside the cube. Recraft V4 produced a slightly more stylized and polished image with superior glass rendering, but failed physics by showing a floating sphere. FLUX.2 [dev] is the winner for its superior prompt adherence and natural 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.”

FLUX.2 [dev]
Recraft V4
50% wins 0% ties 50% wins

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent skin texture and realistic facial details
  • + Accurate motion blur on the passing vehicle in the background
  • + Stronger 'candid' feel with the tighter 50mm-style framing
  • The physics of the bicycle handlebars and cables are slightly confusing
  • Rain is very subtle, almost unnoticeable on the clothing

Recraft V4

  • + Very effective reflections on the wet pavement
  • + Good sense of depth with multiple layers of street activity
  • + Captures the environment well with signs and pedestrians
  • The subject's face is obscured and less detailed
  • The bicycle has structural issues particularly where the frame meets the wheel
  • The rain effect looks like a post-process overlay rather than being integrated

Verdict: FLUX.2 [dev] produces a much more realistic portrait with superior skin textures and a more convincing 'candid camera' aesthetic. Recraft V4 captures the street atmosphere and wet pavement reflections well, but it fails on the finer details of the man's face and the mechanical structure of the bicycle.

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]
Recraft V4
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent adherence to the bead requirement with colorful, distinct beads in the braids.
  • + Very high-quality skin texture showing realistic pores and fresh battle wounds.
  • + Superior background composition with a clear, bright torch providing strong directional lighting.
  • The engraving on the armor looks slightly flat compared to the metal's surface.
  • The 'battle-worn' appearance feels a bit like makeup rather than deep-seated grime.

Recraft V4

  • + Exceptional armor engraving that feels embossed and physically part of the plate.
  • + Better depiction of a 'battle-worn' character with more natural-looking dirt and intensity.
  • + Excellent leather and cloth textures underneath the armor pieces.
  • The beads in the hair are very subtle and blend into the hair color.
  • The lighting is somewhat muddy compared to the requested warm torchlight reflection.

Verdict: FLUX.2 [dev] excels in literal prompt adherence, specifically regarding the beads and the lifelike quality of the eyes and skin. However, Recraft V4 captures the 'battle-worn' paladin aesthetic with more grit and superior armor detailing, making it feel more like a cohesive character portrait. FLUX.2 [dev] is the winner for its clarity, vibrant lighting, and precise execution of all requested 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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent vibrant accents and professional color palette
  • + Good grid layout with integrated photos
  • + Captures a high-end casual dining feel
  • Text is mostly gibberish or misspelled
  • Layout is slightly cluttered with overlapping elements

Recraft V4

  • + Perfectly legible English text and descriptions
  • + True minimalist aesthetic with consistent white backgrounds for food
  • + Strict adherence to the requested grid and sections
  • Minimalist to a fault, looks slightly like a digital ordering app
  • Less 'vibrant' branding compared to Model A

Verdict: Recraft V4 is the clear winner because it produces a functional, legible menu with perfect English and realistic pricing, whereas FLUX.2 [dev] produces gibberish text despite its appealing visual style. Recraft V4 also perfectly followed the request for specific sections (Appetizers, Pizza, Mains) and a clean grid layout.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent typography with high-quality fiery glow effects that match the background.
  • + Well-composed layout that feels like a professional marketing asset.
  • + Strong adherence to the 'fiery' atmosphere requested in the prompt.
  • The burger components are a bit static and vertically aligned rather than 'exploded' or dynamic.
  • The cheese looks somewhat plastic/artificial in its texture.

Recraft V4

  • + Much more dynamic 'exploded' composition with components scattered naturally in 3D space.
  • + Superior photographic texture on the meat and bun, feeling more like real food photography.
  • + Integrated text follows a perspective curve for a more custom design look.
  • Failed to apply the requested 'fiery effect' to the starburst and most of the text.
  • Background is slightly less impactful and simpler compared to the other model.

Verdict: FLUX.2 [dev] produced a more polished and cohesive advertisement that perfectly captured the fiery aesthetic and typography requirements, though the burger itself is a bit rigid. Recraft V4 excelled in the food photography and dynamic composition, but failed to apply the requested glowing fire effects to the text elements. FLUX.2 [dev] is the likely winner for better overall prompt adherence and marketing-ready visual style.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent chalk texture that looks authentically handwritten.
  • + High prompt adherence for the specific cursive title and item descriptions.
  • + Realistic smudging and board texture enhance the cozy café vibe.
  • One minor text smudge artifacts behind the word 'with'.
  • The bottom text line, while legible, is slightly less 'handwritten' in character than the main body.

Recraft V4

  • + Beautiful composition and background environment.
  • + Included charming chalk illustrations for each menu item.
  • + Very clean and readable layout.
  • The font looks digital and uniform rather than authentic chalk handwriting.
  • Failed to provide the 'elegant cursive' requested for the title.
  • The text lacks the natural variation in letter size and slant specified in the prompt.

Verdict: FLUX.2 [dev] followed the stylistic instructions much more closely, delivering a convincing chalk texture and authentic handwriting with natural variations. While Recraft V4 produced a more aesthetically pleasing overall scene with nice illustrations, its text looks like a digital font overlay, failing the primary requirement of the prompt.

The Reversed Rodeo

Text-to-Image

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

FLUX.2 [dev]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent anatomical detail on the horse
  • + High color contrast with vibrant nebula and planet backgrounds
  • + Clean and clear cinematic lighting
  • Failed the spatial reasoning instruction 'horse on top'
  • The composition is a fairly standard 'astronaut riding horse' trope

Recraft V4

  • + Atmospheric use of texture and debris to create a sense of movement
  • + Strong cinematic composition with a dynamic angle
  • + Detailed texture on the astronaut suit and horse hair
  • Failed the spatial reasoning instruction 'horse on top'
  • Low contrast/muted colors make the subject blend slightly into the rocks

Verdict: Both FLUX.2 [dev] and Recraft V4 failed the complex spatial reasoning prompt to place the horse on top of the astronaut, instead defaulting to the common trope of an astronaut riding a horse. FLUX.2 [dev] is slightly preferred for its superior clarity and vibrant color palette, whereas Recraft V4 chose a more cluttered, monochromatic aesthetic that makes the subjects harder to distinguish.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent close-up detail of the capybara's fur and expression.
  • + High-quality rendering of the businesswoman's facial expression and interaction with the phone.
  • + Superior focus and lighting on the taxi interior.
  • The capybara's paws look slightly more like human hands than paws.
  • The taxi cap logo is a generic 'T' rather than specific NYC taxi branding.

Recraft V4

  • + Natural composition showing the full side-profile of the driver and passenger.
  • + Excellent text rendering on the hat ('NYC TAXI').
  • + Atmospheric background with rain effects on the window.
  • The businesswoman's face is slightly blurry and lacks the 'bored' detail requested.
  • The capybara's paws are positioned awkwardly on the steering wheel.

Verdict: FLUX.2 [dev] provides a sharper, more detailed image with better facial expressions, particularly for the businesswoman in the back seat. However, Recraft V4 captures the 'NYC TAXI' branding and the overall scene composition better, including the atmospheric rain. FLUX.2 [dev] is preferred for its technical clarity and character rendering.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent typography rendering with perfect adherence to the requested text.
  • + Very strong composition with a clear hierarchal layout for an invitation.
  • + The combination of thorns and webs creates a high-quality, distinctive border.
  • The parchment texture is limited to the very outer edges.
  • The twisted trees are somewhat obscured by the dark background.

Recraft V4

  • + Highly detailed environment with falling leaves, rain, and numerous bats.
  • + Excellent cinematic lighting on the Jack-o-lantern and forest floor.
  • + Complex and delicate spiderweb border design.
  • The 'scroll banner' is just a simple white tag that lacks a vintage/scroll aesthetic.
  • The text layout is slightly less integrated into the 'poster' feel compared to the other model.
  • One bat in the center looks a bit like a bird/hybrid.

Verdict: FLUX.2 [dev] produced a superior invitation layout with much better typography and a proper scroll element that feels like a professional graphic design. Recraft V4 has more impressive environmental textures and cinematic lighting, but it fails on the specific 'scroll banner' request and has a less cohesive text hierarchy.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent adherence to the 'cartoon' and 'refined texture' stylistic prompts with a smooth 3D aesthetic.
  • + The text is perfectly rendered and follows the requested hierarchy and placement precisely.
  • + Very clean isometric presentation on a well-defined diorama base.
  • The lighting is slightly flat compared to more realistic PBR renderings.
  • The sushi variety is a bit clustered together on the plate.

Recraft V4

  • + Impressive realistic PBR materials, particularly in the fish and the crystalline base.
  • + Creative interpretation of the 'diorama base' as a block of ice.
  • + Sharp, high-clarity image with strong shadows conveying depth.
  • Failed the text hierarchy by placing the flag above the word 'JAPAN'.
  • Missed the 'cartoon scene' stylistic requirement, opting for hyper-realism instead.
  • The text spacing for 'SUSHI' is excessively wide, making it less readable.

Verdict: FLUX.2 [dev] followed the specific instructions for text placement and the 'cartoon 3D' aesthetic much better than Recraft V4. While Recraft V4 achieved superior material realism with its ice-like base and fish textures, it ignored the requested text layout and the specific stylistic direction of the prompt.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent texture on fur and whiskers
  • + Realistic lighting with soft backlighting and God rays
  • + Clean, masterpiece-quality rendering
  • Static composition rather than 'tumbling and chasing'
  • Included an extra rabbit that wasn't' requested

Recraft V4

  • + Captures requested movement of 'chasing' and 'tumbling' perfectly
  • + Includes all requested animals and many butterflies
  • + Dynamic composition with great depth of field
  • The fox's front legs look anatomically distorted while running
  • Some butterflies look slightly like stickers rather than being integrated into the scene

Verdict: Recraft V4 followed the complex action requirements of the prompt much more effectively, showing the animals in motion rather than posing. While FLUX.2 [dev] produced a cleaner, more realistic static portrait, it failed to capture the 'tumbling' and 'chasing' aspect and added an extra animal.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Excellent typography with a classic banner design.
  • + Subtle, realistic texture on the background and logo.
  • + Highly balanced composition that feels unified.
  • The Cloche dome is a bit generic and flat in design.

Recraft V4

  • + Sophisticated, custom-feeling typography that is very stylish.
  • + Better shading and detail on the cloche dome.
  • + Clean vector graphics appearance.
  • Failed to include the requested banner for the 'Est. 1720' text.
  • Background texture is nearly invisible compared to Model A.

Verdict: Both models followed the prompt well, but FLUX.2 [dev] followed the specific request for a banner and a textured background more accurately. Recraft V4 has arguably more beautiful typography, but it missed the banner element and has a more sterile background, making FLUX.2 [dev] the more complete interpretation of the prompt.

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]
Recraft V4

AI Judge Analysis

FLUX.2 [dev]

  • + Features more detailed and visually appealing illustrations
  • + Accurately represents the lunar module and Saturn V silhouettes
  • + Creative interpretation of the astronaut names in a header tag
  • Shows serious text legibility issues and gibberish labels
  • Failed the logical order of steps, creating a confusing grid layout
  • Misplaced step labels like 'Saturn Icon' and 'Descent' (pointing to moon surface)

Recraft V4

  • + Excellent adherence to the 'flat-vector' style and icon consistency
  • + Perfect text rendering and logical chronological flow
  • + Strong adherence to the NASA-inspired color palette
  • Icons are extremely simplified and less representational of the actual Apollo hardware
  • Composition feels a bit sparse with significant empty space

Verdict: Recraft V4 is the winner as it functions as a proper infographic with perfectly legible text and a logical step-by-step layout. While FLUX.2 [dev] produced more sophisticated illustrations, its failure to organize the steps logically and its illegible text labels made it fail the primary goal of the prompt.

Next steps

Explore each model