Black Forest Labs' distilled 9 billion parameter image generation model with sub-second inference and multi-reference support
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.2 [klein] 9B
#13 of 32 in Image Editing
Recraft V4
#11 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [klein] 9B
0.0%
win rate
Ties
0.0%
Recraft V4
100.0%
win rate
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Geometric Composition
Text-to-Image“A glass cube on a wooden table. Inside the cube is a small blue sphere. On top of the cube sits a red book. A green plant is behind the cube, partially visible through the glass. Soft window light from the left.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent adherence to lighting instructions with a clear directional light from the left.
- + Highly realistic glass texture and caustic light reflections on the wooden table.
- + Natural-looking plant integration and depth of field.
- − The glass object is more of a rounded vase shape than a geometric 'cube'.
- − The blue sphere appears to be floating unnaturally without a visible base/support or clear refractive logic.
Recraft V4
- + Perfectly renders the 'cube' geometry as requested.
- + Impressive refractive detail inside the sphere, showing reflections of the room windows.
- + Rich, moody lighting and high-quality textures on the book and table.
- − The plant behind the glass is not as clearly visible through the cube as requested.
- − The sphere occupies a very large portion of the cube's interior, straying slightly from 'small sphere'.
Verdict: Recraft V4 followed the geometric requirements much better by rendering an actual cube, whereas FLUX.2 [klein] 9B generated a rounded glass container. While FLUX.2 handled the lighting and the 'partially visible through glass' plant instruction with more realism, Recraft V4 produced a more striking image with superior refractive details and accurate shape interpretation.
Candid Street Photography
Text-to-Image“A candid street photo of an elderly Japanese man repairing a red bicycle in light rain, reflections on wet pavement, shallow depth of field, 50mm lens, natural skin texture, imperfect framing, motion blur from passing cars, cinematic but realistic, no stylization.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent skin texture and facial detail showing aged characteristics
- + Highly realistic mechanical details of the bicycle
- + Convincing wet pavement reflections and depth of field
- − Lacks the requested motion blur from passing cars
- − Framing feels a bit too balanced and 'perfect' for a candid prompt
Recraft V4
- + Successfully incorporates motion blur and 'imperfect framing' requested in the prompt
- + Captures the atmosphere of a busy street in rain very well
- + Dynamic composition with leading lines and reflections
- − The man's posture and neck anatomy appear slightly distorted and unrealistic
- − Lower level of detail on the man's face and hands compared to the competitor
Verdict: FLUX.2 [klein] 9B produces a much more detailed and anatomically correct image with superior skin textures and mechanical realism. However, Recraft V4 better captures specific atmospheric prompt elements like motion blur and imperfect framing, though it fails on the 'natural skin texture' and anatomical coherence of the subject.
Fantasy Warrior
Text-to-Image“Close portrait of a battle-worn paladin in ornate engraved plate armor, hair braided with small beads, faint scars and dirt on the skin, warm torchlight reflecting off metal, shallow depth of field, bokeh sparks, lifelike eyes, highly detailed texture on leather straps and cloth underlayer.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent detail on the engraved armor and chainmail underlayer.
- + Lifelike eyes with very clear, sharp highlights.
- + Perfect adherence to the beaded hair braids and leather strap textures.
- − The facial scars look a bit like surface paint rather than deep, battle-worn marks.
- − The bokeh sparks appear slightly floating and disconnected from the depth of the scene.
Recraft V4
- + Atmospheric lighting gives a more gritty and 'battle-worn' feel to the character.
- + The hair braids are more complex and feel more natural within the hairstyle.
- + Stronger sense of depth and cinematic atmosphere.
- − The armor engravings are less sharp and detailed compared to Model A.
- − Some minor texture blurring on the skin and leather components.
- − Lower overall sharpness in the facial features.
Verdict: Both models followed the prompt closely, but FLUX.2 [klein] 9B provides superior micro-details in the armor etchings, leather straps, and eyes, creating a very crisp image. Recraft V4 captures the 'battle-worn' mood more effectively through lighting and grit, but it falls slightly behind in technical texture rendering and clarity.
Modern Clean Menu
Text-to-Image“Modern minimalist restaurant menu design, white background with colorful food photos in grid, sections for appetizers/pizza/mains, bold sans-serif fonts, vibrant accents, clean professional layout for casual dining.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Strong professional layout with distinct header and body sections.
- + Vibrant color accents used in horizontal lines and prices.
- + High-quality, realistic photography within the grid frames.
- − Text is mostly gibberish or contains significant spelling errors.
- − Visual contradictions like showing pizzas under headings for 'Mains' and 'Appetizers'.
Recraft V4
- + Perfectly legible English text and coherent menu items.
- + Excellent adherence to minimal food photography with clean isolated backgrounds.
- + Accurate grid layout that corresponds correctly to the section headers.
- − Composition is slightly more rigid and less like a graphic design piece compared to Model A.
- − Lacks a restaurant name or brand header.
Verdict: Recraft V4 is the clear winner as it produces a functional menu with perfectly legible text and items that match their respective categories. While FLUX.2 [klein] 9B creates a more stylistically 'designed' page layout with branding, it fails on basic logic by labeling four different pizzas as 'Appetizers' and 'Mains' and contains unintelligible text.
Magic Burger Explosion: Fiery Photorealism Challenge
Text-to-Image“Ad for 'Magic Burger'. Dynamic, exploded burger with all components (bun, patty, cheese, lettuce, tomato, sauce) suspended in mid-air. Emphasize photorealistic detail and a sense of motion. Dark, fiery background with glowing embers. Integrate text: 'MAGIC BURGER' as a prominent title, 'LIMITED TIME ONLY' as a secondary message, and '€6.99' in a starburst, all rendered with a fiery, glowing effect.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent typography with a cohesive glowing effect
- + Dynamic composition with a strong sense of explosion
- + High-quality, vibrant food rendering
- − The burger structure is a bit repetitive, showing two patties and multiple layers of the same ingredient rather than a clean deconstruction
Recraft V4
- + Natural and realistic 'exploded' physics showing individual ingredients
- + Highly photorealistic texture on the meat and buns
- + Good layout with saucy details
- − The price tag in the top right is small and lacks the requested 'glowing' effect
- − The secondary text is too small to be readable in an ad context
Verdict: FLUX.2 [klein] 9B is the better total package for an advertisement, offering bold, readable typography that perfectly captures the 'fiery' theme and dynamic energy. Recraft V4 excels at the photorealistic deconstruction of the food itself, but its graphic design elements are weak and fail to meet the specific styling requirements for the 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.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent chalk texture on the board surface with realistic smudging
- + Very clear and legible handwriting style
- + Includes the full requested third item based on prompt logic
- − Spelling error in 'Mushrropom'
- − Title is not in 'elegant cursive' as requested, but rather a print style similar to the body text
Recraft V4
- + Perfect spelling on all requested items
- + Higher quality environmental context and background composition
- + Realistic variation in text sizing and spacing
- − Handwriting style is consistent but lacks the 'elegant cursive' requested for the title
- − The board orientation is horizontal whereas 'Model A' captured a more standard menu board verticality
Verdict: Recraft V4 performed significantly better by accurately spelling all menu items and maintaining high visual quality in the cafe background. While FLUX.2 [klein] 9B captured a more realistic chalk 'smudge' texture, the spelling error on the main menu item made it less effective overall.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent clarity and vibrant color palette
- + Highly detailed rendering of planets and galaxies
- + Accurate representation of an astronaut suit
- − Failed the negative constraint; the astronaut is on top of the horse
- − Includes unwanted gibberish text at the bottom
- − The composition feels a bit like a stock photo montage
Recraft V4
- + Strong cinematic atmosphere and lighting
- + Better integration of the subjects into the environment
- + Realistic textures on the rocks and the horse's coat
- − Failed the negative constraint; the astronaut is on top of the horse
- − The astronaut's proportions look slightly awkward relative to the horse
- − Muted color palette compared to the surreal prompt
Verdict: Both FLUX.2 and Recraft V4 failed the specific negative logic instruction to place the 'horse on top', instead both opting for the conventional astronaut-on-horse imagery. Recraft V4 is the better overall image due to its superior cinematic lighting and atmospheric depth, whereas FLUX.2 is marred by nonsensical text and a less cohesive, cluttered space background.
The Capybara Taxi Driver
Text-to-Image“Photorealistic scene inside a yellow New York taxi at night. A capybara is driving, wearing a yellow taxi driver cap and a dark jacket. It has a calm, professional expression and both front paws on the steering wheel. In the back seat sits a human businesswoman in a coat, looking at her phone with a completely normal, bored expression (as if this is just another normal ride). Through the windows you can see the streets of Manhattan at night with blurred lights. Realistic taxi interior, photorealistic, detailed fur and fabric, 35mm lens, night lighting with reflections, shallow depth of field.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent texture on the capybara fur and the businesswoman's coat
- + Perfectly captures the 'bored' expression of the passenger
- + Strong composition with a clear view of both the character and the driver's paws on the wheel
- − Internal car layout is confusing, placing the passenger in the front passenger seat instead of the back seat
- − Text on the hat is nonsensical ('NEW TALA')
Recraft V4
- + Accurately places the passenger in the back seat as requested
- + Clear and accurate text on the driver's cap ('NYC TAXI')
- + Adds realistic weather details like raindrops on the glass and blurred Manhattan architecture
- − The passenger's face is slightly less detailed and more blurred than in Model A
- − The capybara's lower paw is merged strangely with the steering wheel
Verdict: While FLUX.2 [klein] 9B has superior rendering of textures and human expressions, it failed the spatial logic of the prompt by placing the passenger in the front seat. Recraft V4 successfully placed the passenger in the back seat and included accurate text on the hat, while also providing a more convincing NYC atmosphere with rain and tall buildings. Recraft V4 is the winner for better following the complex situational instructions.
The Halloween Invitation
Text-to-Image“Vintage gothic Halloween party invitation. Dark parchment poster, spooky border with webs and thorns, central glowing jack-o-lantern, bats, twisted trees, moody night sky. Add elegant gothic title text saying "Halloween Party Invitation", a small scroll banner saying "You are invited to a night of frights", and event details at the bottom: Date: 30.10.2026 Time: 7pm Location: The Arches, NYC Spooky but polished, cinematic lighting, square format.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent layout with a strong central focus.
- + Superior typography and text rendering.
- + Highly effective interpretation of the thorn and web border.
- − The parchment texture is slightly less pronounced in the background than requested.
Recraft V4
- + Impressive level of fine detail in the foliage and pumpkin texture.
- + Accurate text spelling.
- + Good atmospheric lighting with the moon and mist.
- − The banner for the sub-text is very small and lacks the 'scroll' aesthetic.
- − The composition feels a bit crowded with the number of bats.
- − The secondary text is much harder to read compared to the title.
Verdict: FLUX.2 [klein] 9B followed the design instructions more effectively, particularly regarding the 'scroll banner' and the overall hierarchy of the invitation's text. Recraft V4 produced a very high-quality illustration, but as an invitation, it is less successful due to the tiny sub-text and less cohesive border design.
Isometric Miniature Diorama Scenes
Text-to-Image“Create a clear, 45° top-down isometric miniature 3D cartoon scene of Japan's signature dish: sushi, with soft refined textures, realistic PBR materials, gentle lighting, on a small raised diorama base with minimal garnish and plate. Solid light blue background. At top-center: 'JAPAN' in large bold text, 'SUSHI' below it, small flag icon. Perfectly centered, ultra-clean, high-clarity, square format.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent adherence to the 'cartoon scene' and 'miniature 3D' stylistic prompt.
- + Very clean typography and layout.
- + Smooth, high-quality rendering of textures consistent with PBR materials.
- − The flag icon is incorrect, showing the flag of Yemen rather than Japan.
- − The sushi roll with the salmon on top is a bit structurally confusing.
Recraft V4
- + Perfect accuracy on the Japanese flag icon.
- + Higher realism in the sushi textures specifically.
- + Creative interpretation of the diorama base using a salt or ice texture.
- − Missed the 'cartoon' aspect of the prompt, opting for a photographic style.
- − The typography choice is thin and less impactful than requested.
- − The lighting is harsher than the 'gentle lighting' requested.
Verdict: FLUX.2 [klein] 9B followed the stylistic 'cartoon miniature' instructions much more closely, resulting in a cohesive 3D-render aesthetic, though it failed the flag accuracy. Recraft V4 produced a high-quality photographic image with the correct flag but ignored the specific 'cartoon' and 'bold text' stylistic markers in the prompt. FLUX.2 is preferred for overall adherence to the creative vision despite the flag error.
Adorable Baby Animals in Sunny Meadow
Text-to-Image“Hyper-photorealistic scene of fluffy baby animals—a golden retriever puppy, tabby kitten, baby bunny, and red fox kit—with big expressive eyes and ultra-detailed soft fur, playfully chasing butterflies and tumbling together in a lush wildflower meadow, warm golden sunrise light with god rays and dew sparkles, joyful wholesome vibe, 8K masterpiece.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent lighting and god rays that enhance the 'golden hour' vibe.
- + Very clean textures and expressive faces for the fox and retriever.
- + Vibrant and well-balanced color palette.
- − Completely failed to include the baby bunny requested in the prompt.
- − The kitten's pose and anatomy look slightly awkward behind the other animals.
Recraft V4
- + Successfully included all four requested animals: golden retriever, tabby kitten, bunny, and fox.
- + Captured a more dynamic sense of 'playfully chasing' with motion in the poses.
- + Detailed environment including dew sparkles and diverse flora.
- − The bunny is floating in mid-air in a way that looks physically impossible.
- − The kitten's face appears somewhat flattened and less photorealistic than the other animals.
Verdict: Recraft V4 is the winner because it successfully adhered to the prompt by including the baby bunny, whereas FLUX.2 [klein] 9B missed a core element of the request. While Recraft V4 has some minor anatomical and clipping issues with the floating bunny, its overall composition and completeness make it the superior fulfillment of the instructions.
Vintage Cafe Logo
Text-to-Image“Vintage minimalist restaurant logo for "Caffè Florian", retro cloche dome with steam and "Est. 1720" banner, classic typography, warm brown and cream tones, subtle texture on light background, vector emblem style.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent typography rendering with perfect accents and spacing
- + Accurate representation of the 'Est. 1720' banner as requested
- + Clean vector aesthetic with a pleasant warm color palette
- − The 'subtle texture' on the background is a bit heavy-handed with dark spots
- − The steam coming from the cloche is a bit stylized and thick
Recraft V4
- + Excellent use of negative space and classic serif typography
- + Sophisticated cross-hatch shading on the cloche adds a vintage feel
- + High contrast and very clean vector lines
- − Missed the 'banner' requirement for the establishment date
- − The typography has a minor error with a period added after the name
- − Minimal background texture unlike the prompt request
Verdict: FLUX.2 [klein] 9B followed the prompt more closely by including the specific 'Est. 1720 banner' and a textured background, while maintaining perfect legibility. Recraft V4 produced a more elegant and artistically professional logo, but it missed the banner element and included a stray punctuation mark in the text.
Apollo 11: Journey to Tranquility
Text-to-Image“Create a clean, modern vector infographic poster about the Apollo 11 mission. NASA-inspired palette (navy, white, muted red, light gray). Flat-vector style, crisp lines, consistent iconography, subtle gradients only. Steps (stop at landing): 1. Launch (Saturn Vicon) 2. Earth Orbit (Earth + orbit ring icon) 3. Translunar (trajectory arc icon) 4. Lunar Orbit (Moon + orbit ring icon) 5. Descent (lunar module descending icon) 6. Landing (lunar module on the surface icon) Small supporting elements (minimal text): • Crew strip: three silhouette icons with only last names: Armstrong, Aldrin, Collins. • Landing site marker: Moon pin labeled "Tranquility" only. Layout constraints: generous margins, large readable labels, clean background with subtle stars. Vector-only, print-poster look, high resolution.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Effective use of the requested NASA-inspired color palette
- + Good vector illustration style for the Saturn V and Lunar Module
- + High visual impact and balanced composition
- − Numerous spelling errors in text labels (e.g., AOILLO, EARDHT, LANDINING)
- − The logical flow of the steps is confusing and disconnected
Recraft V4
- + Perfect text rendering with zero spelling errors
- + Clean, consistent iconography that follows the prompt instructions exactly
- + Excellent use of negative space and clear logical flow of mission steps
- − Icon layout is a bit sparse compared to a typical 'poster' format
- − The translunar icon is slightly abstract compared to the others
Verdict: Recraft V4 is the clear winner as it successfully rendered all text labels without errors and followed the specific instructions for each of the six mission steps. While FLUX.2 [klein] 9B has a nice aesthetic, its catastrophic spelling failures and illogical flow make it ineffective as an infographic.
Explore each model
Recraft's latest text-to-image generation model with high-quality output, supporting various aspect ratios and custom color palettes