Black Forest Labs' compact, open-source image generation model with sub-second inference, optimized for production and near real-time applications with multi-reference support
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.2 [klein] 4B
22.3
arena score
#32 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 [klein] 4B
0.0%
win rate
Ties
0.0%
Recraft V4
100.0%
win rate
0.0%
0.0% ties
100.0%
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent photographic realism with natural window light
- + Perfect composition with a clean, modern aesthetic
- + Accurately represents the transparency of the glass cube
- − The plant visibility through the glass is a bit blurred and less distinct
Recraft V4
- + Strong internal reflections and refraction within the sphere
- + Very clear visibility of the plant through the glass cube
- + Sharp details on the texture of the red book
- − The blue sphere appears to be magically floating rather than resting on the bottom
- − The glass cube looks more like a solid block than a hollow container
Verdict: FLUX.2 [klein] 4B produces a much more natural and believable scene with superior lighting and realistic physics. While Recraft V4 captures the prompt elements well, the floating sphere and the density of the glass make the image feel less grounded in reality compared to the high-quality photographic output of FLUX.2.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent depiction of rain droplets and pavement reflections.
- + Superior rendering of facial details and skin texture.
- + The bicycle design, while slightly warped, fits the 'mamachari' aesthetic common in Japan.
- − The cars in the background are static rather than having the requested motion blur.
- − The man appears to be sitting on or hovering over the bike in an unnatural posture for 'repairing'.
- − The bicycle frame has structural inconsistencies around the pedals and chain.
Recraft V4
- + Successfully incorporates motion blur from a passing car as requested.
- + The 'imperfect framing' feels more authentic to a candid street photo.
- + Stronger sense of environment with pedestrians and atmospheric depth.
- − The bicycle is missing its rear half entirely, which is a major anatomical error.
- − The subject's hands and the interaction with the bike are blurry and poorly defined.
- − The rain effect is less convincing and looks more like a post-process filter than Model A.
Verdict: FLUX.2 [klein] 4B produces a much higher quality portrait with impressive skin textures and realistic rain, but it fails to include the requested motion blur. Recraft V4 captures the 'candid' feel and motion blur perfectly, but the image suffers from a significant structural failure where half of the bicycle is missing. FLUX.2 is the preferred model for its overall visual coherence and superior detail.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Extremely clean and clear facial features
- + Ornate and symmetrical engraving patterns on the armor
- + Vibrant colors and high-contrast lighting
- − Lighting feels a bit artificial and 'studio-like'
- − Skin texture looks relatively clean for a 'battle-worn' character
- − The braids are very neat despite the combat theme
Recraft V4
- + Excellent interpretation of 'battle-worn' with grit and intense expressions
- + Naturalistic lighting from the torches with realistic shadows
- + Highly detailed texture on the leather straps and beads
- − Slightly lower clarity on the fine facial skin texture compared to Model A
- − Composition is a bit tighter, cutting off some armor detail
Verdict: Both models followed the prompt very well, but Recraft V4 captured the 'battle-worn' atmosphere more convincingly with grittier skin and realistic lighting. FLUX.2 [klein] 4B produced a very clean and aesthetically pleasing image, but it feels more like a cosplayer in a studio than a warrior on a battlefield.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + High visual appeal in food photography
- + Includes a logo and sophisticated layout elements
- + Strong use of colorful accents
- − Nonsensical text and spelling errors in headings
- − The grid layout feels a bit cluttered and overwhelming for a minimalist theme
- − Inconsistent font rendering in smaller body text
Recraft V4
- + Excellent text legibility and accuracy
- + Perfect adherence to requested menu categories
- + Clean, professional minimalist design with consistent lighting
- − Standard stock-photo appearance lacks the artistic depth of some modern branding
- − Composition is very safe and repetitive
Verdict: Recraft V4 is the clear winner as it produced an actual functional menu with perfectly legible text and the exact categories requested (Appetizers, Pizza, Mains). While FLUX.2 [klein] 4B has more vibrant and artistic food photography, it failed on basic text rendering and prompt-specific sections.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Strong fiery atmosphere with excellent lighting on the burger.
- + Correct price starburst implementation with high resolution textures.
- + Good sense of motion and vibrant colors.
- − Major text failure with scrambled, unreadable title characters.
- − Failed the 'exploded burger' requirement, showing a fully assembled burger instead.
Recraft V4
- + Perfectly captured the 'exploded burger' layout with components suspended in mid-air.
- + Exceptional text rendering with no spelling errors.
- + Realistic food photography style with crumbs and sauce details.
- − The background is less 'fiery' and dark compared to Model A.
- − The starburst for the price looks a bit more like a generic sticker than a glowing effect.
Verdict: Recraft V4 is the clear winner as it followed every complex instruction, including the 'exploded' layout and all specific text strings without error. FLUX.2 [klein] 4B produced a high-quality image, but it failed to follow the structural request for an exploded view and produced incomprehensible gibberish for the main title.
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 [klein] 4B
Recraft V4
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent authentic chalk texture with dust trails and smudge marks.
- + Strong cursive handwriting style used for labels.
- + Accurately captured the dates and currency values.
- − Numerous spelling errors including 'Musheram', 'Ootrpous', and 'Brawn'.
- − The title has a spacing error in 'S PECIALS'.
Recraft V4
- + Perfect spelling for every item on the menu.
- + Includes creative chalk illustrations matching the menu items.
- + Composition includes the 'cozy café' atmosphere requested in the prompt.
- − The text appears too clean and uniform, resembling a digital font rather than authentic handwriting.
- − The title font is a simple sans-serif rather than the requested elegant cursive.
Verdict: Recraft V4 wins on technical accuracy, providing perfect spelling and a complete café scene with charming illustrations. While FLUX.2 [klein] 4B has a much more realistic chalk texture and hand-written feel, its frequent and glaring spelling mistakes make the text-heavy prompt a failure.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.2 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Crisp and clear high-resolution details on the space suit and horse's mane.
- + Vibrant and colorful space background with a clear view of the earth's curvature.
- − Completely failed the negative constraint to have the horse on top.
- − The composition feels a bit like a basic cutout pasted over a background.
Recraft V4
- + Cinematic lighting and atmosphere with a more dynamic, dusty environment.
- + Better integration between the subject and the surroundings through the use of shadows and debris.
- − Completely failed the negative constraint to have the horse on top.
- − The horse's front legs exhibit anatomical glitches and unnatural bending.
Verdict: Both models failed the specific prompt instruction to place the horse on top of the astronaut, providing a standard 'astronaut on horse' image instead. FLUX.2 [klein] 4B is preferred as a general image for its superior clarity and anatomical correctness, whereas Recraft V4 has significant artifacts in the horse's legs despite a more cinematic atmosphere.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent facial features on both characters
- + Highly realistic textures on the capybara fur and jacket fabric
- + Superior composition using the rear-view perspective requested by the prompt
- − The capybara's paws are rendered a bit awkwardly around the steering wheel
Recraft V4
- + Accurate text on the taxi driver cap
- + Good environmental storytelling with rain on the window
- − The passenger's face is somewhat blurry and lacks detail
- − The side-view perspective makes the interior feel less immersive than the prompt implied
- − The capybara's paws look more like human hands/fingers which is disturbing
Verdict: FLUX.2 [klein] 4B produces a much more convincing and high-quality image, with a professional-looking cab interior and a passenger who perfectly matches the bored expression described. Recraft V4 captures the text and the outdoor environment well, but the anatomy of the capybara's hands and the lower quality of the human face make it less successful overall.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Strong composition with a clear central focus.
- + Accurate thorns and cobwebs in a well-defined border.
- + Vibrant cinematic lighting on the jack-o-lantern.
- − Significant spelling errors in the title text (e.g., 'Hallbwom').
- − Missing the 'Time' entry and contains typos in 'Location' and the date.
Recraft V4
- + Perfect text rendering for all requested fields including names, dates, and locations.
- + Highly detailed and atmospheric background with leaves and numerous bats.
- + Excellent adherence to the gothic font and scroll request.
- − The cobweb border is a bit repetitive and less 'vintage' than requested.
- − The 'scroll banner' is more of a modern tag than a classic parchment scroll.
Verdict: Recraft V4 is the clear winner because it correctly rendered all the requested text with zero typos, which is essential for an invitation. While FLUX.2 [klein] 4B had a slightly more cohesive vintage aesthetic, its failure to spell 'Halloween' or include the full event details makes it unusable as a party 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.2 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent 3D miniature/cartoon aesthetic with clean PBR textures
- + Precise shadows and lighting that give a realistic small-scale feel
- + Text is large and bold as requested
- − Misses the small flag icon for Japan, showing a generic striped bar instead
- − The word 'SUSHI' is slightly cut off or strangely aligned with the icon
Recraft V4
- + Perfect text rendering and inclusion of the Japanese flag icon
- + Accurately represents the 'raised diorama base' with a creative ice or salt block effect
- + Includes a wider variety of sushi items
- − Lighting is harsher and less 'gentle' than requested
- − Leans more towards realistic photography than the requested '3D cartoon scene'
Verdict: Both models followed the prompt well, but they excelled in different areas. FLUX.2 [klein] 4B better captured the 3D isometric cartoon aesthetic and gentle lighting, but Recraft V4 was superior in text accuracy, iconography, and interpreting the 'diorama base' requirement. Recraft V4 is the likely winner for its perfect adherence to the text and flag details while still maintaining high visual clarity.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent fur texture and lighting on the animals
- + Clearer, more vibrant god rays and atmospheric effects
- + Strong expressive eyes as requested
- − Failed to include the baby bunny
- − Generated two kittens instead of one
- − Anatomical issues with the animals' paws and legs
Recraft V4
- + Successfully included all four requested animals (dog, cat, bunny, fox)
- + Ccaptured the dynamic 'tumbling' and chasing movement better
- + Better depth of field with butterflies in the foreground
- − The fox's front legs are unnaturally long and thin
- − Slightly less 'masterpiece' finish compared to the lighting in image A
Verdict: Recraft V4 is the winner because it successfully followed the complex prompt by including all four specific animals, whereas FLUX.2 [klein] 4B missed the bunny and duplicated the kitten. Recraft V4 also captured a more energetic and cohesive scene that matched the 'tumbling' and 'chasing' description, despite some minor anatomical awkwardness in the fox's legs.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Clean vector emblem style
- + Well-executed textured background
- + Clear and legible established date
- − Misspelled the name as 'Caffè Flaxtion'
- − Redundant repetition of 'Est. 1720' text
Recraft V4
- + Correctly spelled 'Caffè Florian'
- + Excellent vintage typography and cross-hatch shading on the cloche
- + Compact and balanced composition
- − Failed to include a 'banner' for the date as requested
- − Minimal steam detail compared to the prompt's focus
Verdict: Recraft V4 is the clear winner as it correctly spelled the brand name 'Caffè Florian' and followed the stylistic cues for vintage typography. While FLUX.2 [klein] 4B followed the 'banner' instruction better, it suffered from a major spelling error and redundant text elements.
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 [klein] 4B
Recraft V4
AI Judge Analysis
FLUX.2 [klein] 4B
- + Successfully captures the requested color palette
- + Good vector illustration style for the Saturn V and Lunar Module
- − Significant spelling errors in the title and labels
- − Poor icon consistency and confusing layout flow
- − Failed to include all six requested steps distinctly
Recraft V4
- + Excellent text rendering with no spelling errors
- + Followed all six specific steps in chronological order
- + Clean, consistent icon set that matches the flat-vector style perfectly
- − The composition has a bit of unnecessary empty space in the bottom left
- − Icon for Saturn V is very simplified compared to model A
Verdict: Recraft V4 clearly wins this challenge by adhering to every specific step requested in the prompt with perfect spelling and high-quality, consistent iconography. While FLUX.2 [klein] 4B has a nice artistic rendering of the spacecraft, its text is illegible and it fails to logically organize the mission steps as requested.
Next steps
Explore each model
Recraft's latest text-to-image generation model with high-quality output, supporting various aspect ratios and custom color palettes