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
#32 of 62 in Text-to-Image
Qwen Image
#37 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [klein] 4B
0.0%
win rate
Ties
0.0%
Qwen Image
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] 4B
- + Excellent texture on the red book cover
- + Sophisticated glass refraction and realistic reflections
- + Highly convincing wooden table texture
- − The glass cube looks more like a 5-sided box with an open top where the book sits
- − The blue sphere is slightly off-center
Qwen Image
- + Perfect adherence to all spatial relationships in the prompt
- + Clean and vibrant colors
- + Accurate representation of the glass cube's structure
- − The textures look slightly more digital/artificial than Model A
- − The lighting on the sphere is very uniform
Verdict: Both models followed the prompt instructions perfectly, including the specific lighting and object placement. FLUX.2 [klein] 4B produced a more photorealistic image with superior textures, particularly the grain of the wood and the fabric of the book, whereas Qwen Image delivered a very clean, structured composition that clearly defined every element of the cube.
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] 4B
- + Excellent skin texture and facial detail on the subject.
- + Strong adherence to the 'imperfect framing' prompt giving it a realistic street photography feel.
- + Vibrant but realistic color palette and lighting.
- − The structural anatomy of the bicycle is slightly nonsensical around the chain guard and frame.
- − Lack of visible 'motion blur' on the passing cars as requested.
Qwen Image
- + Successfully captured the motion blur of the passing car as requested.
- + Better overall bicycle anatomy and structural coherence.
- + Great reflection work on the wet pavement.
- − The facial texture is somewhat smooth and lacks the 'natural skin texture' detail of the competitor.
- − The rain effect is less convincing and looks more like static vertical lines.
Verdict: FLUX.2 [klein] 4B produces a much more immersive and high-detail portrait with a convincing street photography aesthetic, though it fails to include the requested motion blur. Qwen Image adheres more closely to all technical prompt requirements, including motion blur, but the final image quality and skin rendering feel less realistic and more like a standard AI generation.
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] 4B
- + Excellent realism in skin texture and facial features
- + Highly detailed and consistent engraving on the plate armor
- + Subtle and natural integration of scars and dirt
- − Lighting is a bit flat across the face for being lit by torches
- − Braids are quite simple with few beads
Qwen Image
- + Stronger adherence to the hair beads request with colorful variety
- + More dramatic lighting and atmosphere with better depth of field
- + Detailed leather straps and cloth underlayer textures
- − Facial scars look a bit painted on rather than natural
- − Armor engravings are slightly less sharp than the other model's
Verdict: FLUX.2 [klein] 4B produces a much more realistic human face and believable metal textures, making it feel like a genuine photograph. Qwen Image offers a more cinematic and stylized interpretation with more emphasis on the colorful braids and dramatic torchlight, though the scars appear less lifelike. FLUX.2 is preferred for its technical execution of textures and facial realism.
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] 4B
- + Features more sophisticated food photography with appetizing lighting.
- + Provides a complex layout with distinct columns and pricing structures.
- + The typography feels more 'modern minimalist' in its spacing and weight.
- − Nonsense filler text and misspelled headers like 'MDAINS' and 'FRAZZA'.
- − The grid of images is slightly inconsistent in alignment.
Qwen Image
- + Stronger adherence to the 'grid' layout requested in the prompt.
- + Recognizable section headers like 'Appetizers' and 'Pizza' with cleaner spelling.
- + Bold, high-contrast typography that is very legible.
- − Food photos look more generic and 'clip-art' style compared to image A.
- − Visual balance is slightly off with a large empty coral-colored block in the center.
- − Text-to-image artifacts on the sub-text lines.
Verdict: Both models followed the prompt well, but FLUX.2 [klein] 4B produced much more professional and realistic food photography, making it feel like a genuine restaurant menu despite the spelling errors. Qwen Image adhered better to the grid layout and produced more legible header text, but the overall aesthetic feels slightly more like a flyer than a high-end menu design.
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] 4B
- + Excellent photorealistic lighting and texture on the bun and meat.
- + Accurate starburst asset for the price tag.
- + Good integration of sparks and motion blur.
- − Failed the main text rendering with scrambled, overlapping letters.
- − Did not follow the 'exploded' instruction, showing a mostly assembled burger.
Qwen Image
- + Perfect text rendering for all requested strings.
- + Accurately depicts an 'exploded' burger with suspended components.
- + Excellent fiery background and glowing effect on the logo.
- − Individual food pieces like the floating lettuce and tomato look slightly plastic or artificial.
- − The starburst shape is a bit simplified compared to the rest of the image.
Verdict: While FLUX.2 [klein] 4B has superior photorealistic textures, it failed completely on the primary text and the 'exploded' composition. Qwen Image successfully followed every technical instruction in the prompt, including complex text rendering and the specific suspended-component layout, making it the much more effective advertisement.
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] 4B
- + Excellent realistic chalk texture with smudge marks and dust.
- + Maintains consistent cursive handwriting style throughout.
- + Accurately represents the requested date and prices.
- − Numerous spelling errors in the menu items (e.g., 'Truffel Musheram', 'Ootrpous', 'Buter').
- − Title has a spacing error with a double 'S' in 'SPECIALS'.
Qwen Image
- + Perfect spelling of all menu items.
- + Complete and logical layout within a café environment.
- + Very legible text compared to a handwritten style.
- − The year is incorrectly rendered as '20026' instead of '2026'.
- − The text looks more like a digital font or vector graphic than natural chalk handwriting.
- − Lack of chalk texture variation requested in the prompt.
Verdict: FLUX.2 [klein] 4B captures the requested chalk texture and artistic 'handwritten' aesthetic beautifully, but suffers from significant spelling issues and poor character spacing. Qwen Image provides perfect spelling and a clearer environmental context, but its failure to render the correct year and its reliance on a font-like appearance misses the core 'natural chalk variations' requested in the prompt. FLUX.2 is the winner for better adhering to the specific stylistic constraints despite the literal typos.
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] 4B
- + Excellent anatomical rendering of both the astronaut and the horse.
- + Rich, cinematic lighting with a highly detailed starfield background.
- + Effective use of shadows on the planetary surface below adds to the depth.
- − Failed the negative constraint to put the horse on top of the astronaut.
Qwen Image
- + Clean composition with a nice secondary planet for visual balance.
- + Good texture work on the spacesuit and horse's coat.
- − Failed the negative constraint to put the horse on top of the astronaut.
- − The horse's legs have some anatomical awkwardness in the posing.
Verdict: Both FLUX.2 [klein] 4B and Qwen Image failed the specific spatial reasoning challenge to place the 'horse on top' of the astronaut, both reverting to the common 'astronaut on a horse' trope. FLUX.2 [klein] 4B is the winner due to significantly higher visual fidelity, better lighting, and more realistic anatomy compared to the slightly more generic feel of Qwen Image.
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] 4B
- + Excellent photorealism with convincing textures on both the capybara fur and the human's clothing.
- + Perfect adherence to the perspective, capturing the scene from 'inside' the taxi as requested.
- + Captures the bored expression of the woman and the professional demeanor of the capybara effectively.
- − The steering wheel and hands are slightly blurred and lack some structural definition.
Qwen Image
- + Sharp lighting and high-quality rendering of the capybara's head and features.
- + Includes a realistic 'taxi' sign on the roof seen through the perspective.
- − The steering wheel is held by primate-like hands rather than capybara paws.
- − The perspective is more from the passenger window rather than truly 'inside' the rear seating area as specified.
- − The woman appears to be in the front passenger seat or middle, rather than the back seat.
Verdict: FLUX.2 [klein] 4B is the clear winner for its superior composition, which feels more like a cinematic shot from inside the taxi. While Qwen Image has good lighting, it fails the prompt by giving the capybara primate hands and placing the businesswoman in the front passenger area rather than the back seat.
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] 4B
- + Excellent visual style with cinematic lighting and a dense atmospheric background.
- + High-quality rendering of the jack-o-lantern and twisted trees.
- + Comprehensive inclusion of all border elements like thorns and webs.
- − Major spelling errors in the main title ('Hallbwom Party niisation').
- − Several typos in the date and location text ('206' instead of '2026', 'Loeciton').
Qwen Image
- + Accurate rendering of the date, time, and location information with no typos.
- + Stronger parchment aesthetic for the overall poster texture.
- + Good layout that fits all elements into a cohesive frame.
- − The main title has a slight overlap error ('Halle Party Invitation').
- − The graphical style is a bit flatter compared to the cinematic quality of the competitor.
Verdict: While FLUX.2 [klein] 4B produces a much more visually stunning and atmospheric image, its text rendering is poor with significant spelling errors throughout. Qwen Image successfully captures almost all the required text accurately, including the complex date and location details, making it the more functional invitation despite slightly less sophisticated lighting.
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] 4B
- + Excellent PBR material rendering on the salmon and plate textures.
- + Clean, professional typography and layout.
- − Missed the 'small raised diorama base' requirement, using just a plate.
- − The flag icon is incorrect, resembling the flag of Austria rather than Japan.
Qwen Image
- + Perfectly followed the diorama base and isometric 3D cartoon style instructions.
- + Accurate Japanese flag rendering in both the icon and the scene.
- + Stronger adherence to the 'miniature' aesthetic requested.
- − The white text on a light blue background has slightly lower contrast than black text.
- − Maki roll rice texture is slightly repetitive.
Verdict: Qwen Image is the clear winner as it successfully incorporated every element of the prompt, including the specific isometric diorama base and the correct flag of Japan. While FLUX.2 exhibited high-quality realistic textures, it failed on the core structural requirement of the diorama and produced the wrong national 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.”
AI Judge Analysis
FLUX.2 [klein] 4B
- + Excellent fur texture and lighting on the fox and golden retriever.
- + Stronger '8K masterpiece' feel with high-frequency details in the grass and flowers.
- + Dynamic composition with animals interacting more actively with the environment.
- − Missed the baby bunny prompt requirement, including two kittens instead.
- − The puppy's proportions are slightly large compared to the other animals.
Qwen Image
- + Followed the prompt perfectly including all four specific animals: puppy, kitten, bunny, and fox.
- + Excellent rendering of the 'dew sparkles' and soft bokeh effect.
- + Genuinely cute, expressive faces on all creatures.
- − Slightly less realistic, leaning more towards a hyper-clean digital art style.
- − The golden retriever's head shape is a bit overly symmetrical and flat.
Verdict: While FLUX.2 [klein] 4B produces a slightly more photorealistic texture with impressive lighting, it failed to include the baby bunny, instead generating two kittens. Qwen Image followed the complex prompt precisely, successfully including the puppy, kitten, bunny, and fox kit while maintaining a high level of visual appeal and detail.
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] 4B
- + Clean vector aesthetic with crisp lines.
- + Excellent background texture that feels like vintage paper.
- + Professional and balanced layout.
- − Failed to spell the restaurant name correctly, outputting 'FLAXTION'.
- − Redundant 'Est. 1720' text appearing twice.
Qwen Image
- + Captured the spelling of the name better than the competitor.
- + Good warm brown and cream color palette.
- + Clearer steam illustration relative to the cloche.
- − Confusing typography with overlapping letters and inconsistent sizes.
- − The spelling 'LoraiAN' is still incorrect despite being closer.
- − Muddier icon details compared to the crispness of Model A.
Verdict: While both models failed to perfectly render the name 'Caffè Florian', FLUX.2 [klein] 4B produced a much more professional and aesthetically pleasing vector logo design. Qwen Image attempted a more creative layout, but the overlapping text and distorted letters resulted in a messy, unusable composition.
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] 4B
- + Excellent adherence to the muted color palette.
- + Clean vector illustration style with sophisticated gradients.
- + Stronger individual icon designs for the planet and lunar module.
- − Poor text rendering with many nonsensical words.
- − Confusing layout that does not clearly follow a chronological flow.
- − Inaccurate rocket depiction that looks more like a modern capsule-topped rocket than a Saturn V.
Qwen Image
- + Text is much more readable and coherent, even including the astronaut names.
- + Better 'process' flow with arrows and numbered steps.
- + Visualizes the moon landing surface clearly at the bottom.
- − Included the prompt instruction '(Stop at landing)' inside the actual image text.
- − Graphic style is a bit more simplistic and 'clip-art' like compared to the other model.
- − The Saturn V icon is very stylized and loses its iconic silhouette.
Verdict: Qwen Image is the preferred choice because it successfully creates an actual infographic with a logical flow, numbered steps, and relatively clear text, despite including meta-text from the prompt. FLUX.2 [klein] 4B has much higher artistic quality and a more professional color palette, but it fails as an infographic due to illegible text and a disorganized layout.
Explore each model
Alibaba's Qwen image model