Black Forest Labs' 12-billion parameter flow transformer for high-quality text-to-image generation, suitable for personal and commercial use with streaming support
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 [dev]
#16 of 62 in Text-to-Image
Qwen Image
#40 of 62 in Text-to-Image
Where the votes landed
FLUX.1 [dev]
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.1 [dev]
- + Excellent high-resolution textures on the red book cover and wood grain.
- + Sophisticated lighting with realistic caustic reflections from the glass and sphere.
- + Striking, large-scale composition of the primary subjects.
- − The sphere appears more like a large marble than a 'small sphere'.
- − The plant behind the cube is very blurry, making it hard to see 'through the glass' as requested.
Qwen Image
- + Better adherence to the 'small sphere' relative scale within the cube.
- + The plant is clearly visible through the glass panels of the cube.
- + Accurate spatial arrangement and soft side-lighting.
- − Texture quality is slightly flat compared to Image A.
- − The reflective base of the cube looks more like a mirror than transparent glass.
Verdict: Both models followed the prompt instructions near-perfectly regarding the arrangement of objects. FLUX.1 [dev] produced a more photorealistic image with superior textures and light behavior, while Qwen Image followed the size specifications for the 'small sphere' more accurately and made the plant clearly visible through the transparency 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.1 [dev]
- + Excellent texture on the skin and clothing
- + Realistic lighting and wet pavement reflections
- + Accurate shallow depth of field with a 50mm look
- − The man is holding the bicycle rather than repairing it
- − Motion blur on the background car is subtle but static
Qwen Image
- + Clearer interaction with the bicycle suggestive of a 'repair' action
- + Good use of wet pavement reflections
- + Adheres well to the imperfect framing request
- − Noticeable anatomical issues with the hands
- − Low-quality background car with zero motion blur as requested
- − Lower overall resolution and clarity compared to Model A
Verdict: FLUX.1 [dev] produces a significantly more realistic image with superior textures and lighting, though it interprets 'repairing' as simply holding the bike. Qwen Image attempts a more active repair pose but suffers from distorted hands and a sharp background car that ignores the 'motion blur' prompt. FLUX.1 [dev] is the clear winner for its cinematic quality and realistic details.
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.1 [dev]
- + Extremely realistic skin texture and lifelike eyes
- + High quality cinematic lighting and shallow depth of field effects
- + Strong focus on the face as requested in the portrait prompt
- − Missed the request for beads in the hair
- − Armor is simple and lacks the specific 'ornate engraved' detail requested
- − Character looks more like a fashion model than 'battle-worn'
Qwen Image
- + Strong adherence to details like beads in hair, ornate engravings, and leather straps
- + Better interpretation of 'battle-worn' with visible scars and blood
- + Includes the torchlight source and sparks directly in the composition
- − The sparks have a slightly artificial, stamped-on appearance
- − The iris and eyeball textures are less realistic than Model A
- − Composition is slightly less 'close' than a standard close portrait
Verdict: While FLUX.1 [dev] produced a more photorealistic and aesthetically pleasing face, Qwen Image followed the complex prompt requirements much more accurately. Qwen Image successfully included the beads, ornate engravings, and specific leather strap textures that FLUX.1 [dev] omitted, making it the better choice for this specific prompt.
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.1 [dev]
- + Excellent typographic hierarchy with professional weighting.
- + Highly realistic food photography with clean shadows.
- + Strong adherence to the 'modern minimalist' aesthetic with white space.
- − Failed the request for a 'grid' of food photos, placing only a few items decoatively.
- − Missing a dedicated 'Pizza' section header.
Qwen Image
- + Successfully implemented the requested grid layout for food photos.
- + Includes the requested sections for Appetizers and Pizza/Mains.
- + Vibrant color palette that makes the menu feel approachable for casual dining.
- − Text rendering is poor with significant artifacts and gibberish characters.
- − The balance between images and text is slightly skewed, making it less practical as a functional menu.
Verdict: FLUX.1 [dev] produces a much more professional and realistic graphic design piece with superior typography and image quality, though it ignored the 'grid' instruction. Qwen Image captured the specific structural requirements of the prompt like the grid and specific section titles but failed significantly on the execution of readable text and visual clarity. FLUX.1 [dev] is the better choice for a usable design template despite the minor layout deviation.
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.1 [dev]
- + Excellent photorealistic texture on the meat and buns
- + Clean vertical composition creating a levitation effect
- − Failed to include the primary title 'MAGIC BURGER'
- − Missing the starburst element for the price
- − The 'exploded' effect is just stacked layers rather than a dynamic motion burst
Qwen Image
- + Perfect adherence to text requirements, including the starburst and fiery glowing title
- + Highly dynamic composition with ingredients flying outward, conveying motion
- + Strong background integration with glowing embers and fire
- − Some ingredients like the ketchup dollops look slightly cartoonish compared to the photorealistic burger patty
Verdict: Qwen Image is the clear vinner as it followed all complex text and layout instructions, including the specific 'MAGIC BURGER' title and starburst price tag which FLUX.1 [dev] omitted. While FLUX.1 [dev] produced a very clean and realistic food stack, it failed to capture the 'exploded' motion and essential brand elements requested in the prompt.
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.1 [dev]
- + Excellent text rendering with perfect spelling for all prompt items.
- + Realistic representation of chalk texture on a dark blackboard.
- + High-quality wood grain texture on the menu frame.
- − The cursive title is fairly standard rather than 'elegant'.
- − The bottom text contains minor typos ('drish', 'usfor', 'aur').
Qwen Image
- + Successfully captured the requested slight slant in the handwriting style.
- + Applied a convincing chalk-dust smudge effect to the background of the board.
- + Good composition that feels integrated into a café table setting.
- − Failed the date prompt by writing '20026' instead of '2026'.
- − The text lacks the requested chalk texture, appearing more like a digital pen or marker.
- − Bottom right corner contains illegible text artifacts.
Verdict: FLUX.1 [dev] followed the text prompts much more accurately, correctly spelling the food items and the specified year, whereas Qwen Image failed on the date. While Qwen Image captured the 'slant' requested in the handwriting, FLUX.1 [dev] provided a much more realistic chalk-on-blackboard texture.
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.1 [dev]
- + Excellent cinematic lighting with a soft glow on the horse's mane.
- + Smooth, high-quality textures on the space suit and saddle.
- + Dynamic composition with a strong diagonal line against the horizon.
- − The horse's front left leg has anatomical issues, appearing to merge with the neck.
- − The hind legs have slightly distorted hoof shapes.
Qwen Image
- + Realistic anatomical proportions for the horse.
- + Detailed space suit design featuring patches and realistic tubing.
- + Includes additional celestial bodies like a moon to enhance the space setting.
- − The harness/bridle lines are messy and logically inconsistent near the astronaut's hands.
- − The lighting is somewhat flat compared to the other model.
Verdict: Both models failed the negative constraint to have the horse on top of the astronaut, instead opting for the standard 'astronaut riding a horse' interpretation. FLUX.1 [dev] is the preferred choice due to its superior cinematic lighting and artistic polish, whereas Qwen Image suffered from messy line work in the horse's tack.
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.1 [dev]
- + Excellent photographic quality and lighting on the capybara's fur.
- + Accurately places the passenger in the back seat as requested.
- − The passenger is sitting in the center of the frame alongside the driver instead of clearly in the back seat row.
- − Includes a second steering wheel in front of the passenger.
- − The capybara's paws look more like bird claws.
Qwen Image
- + Perfect composition with the passenger clearly in the back seat.
- + Highly realistic taxi driver jacket and cap detail.
- + The capybara's hands are blended more realistically with the steering wheel.
- − The text on the taxi sign says 'YOXI' instead of 'TAXI'.
- − Slightly less 'professional' expression on the capybara compared to Model A.
Verdict: Qwen Image followed the spatial instructions much better, placing the passenger correctly in the back seat and maintaining a single steering wheel, whereas FLUX.1 [dev] suffered from a serious composition error with a dual-steering wheel layout. Qwen Image also captured the 'bored' expression of the businesswoman perfectly while maintaining a high level of photorealism.
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.1 [dev]
- + Strong cinematic lighting and vibrant colors.
- + Cohesive and elegant thorn border.
- + Clean layout with a central focal point.
- − Major text errors including 'Halloween Rant' and several typos in the event details.
- − Missing the cobweb elements and the dark parchment texture requested.
Qwen Image
- + Better adherence to the 'dark parchment' and 'web' style requirements.
- + Higher overall text accuracy, specifically with the date, time, and location.
- + More atmospheric and spooky moody night sky.
- − The main title 'Halloween Party Invitation' has overlapping characters and minor misspellings.
- − Composition feels a bit more cramped compared to the other model.
Verdict: Qwen Image followed the prompt much more closely by including the specific parchment texture and cobweb details that FLUX.1 [dev] omitted. While FLUX.1 [dev] has superior lighting and a cleaner border, its failure to spell the primary title correctly ('Halloween Rant') and the omission of key stylistic requests make Qwen Image the more functional 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.”
AI Judge Analysis
FLUX.1 [dev]
- + Excellent PBR material rendering on the salmon and base
- + Very clean and minimal diorama aesthetic
- + Beautiful soft lighting and shallow depth of field
- − Failed to render the word 'SUSHI' correctly, adding 'SUSH CATON' and a random kanji
- − The flag icon is integrated into a graphic rather than being a clean icon element
Qwen Image
- + Perfect adherence to text requirements with 'JAPAN' and 'SUSHI' rendered correctly
- + Excellent interpretation of the 'cartoon' style while maintaining 3D depth
- + Includes the flag icon exactly as requested in the top center
- − The sushi models are slightly more simplistic and less realistic in texture compared to Model A
- − Perspective on the text at the top is flat compared to the 3D scene
Verdict: While FLUX.1 [dev] produced a more sophisticated and realistic texture/lighting setup, it failed significantly on the text rendering. Qwen Image captured the intended '3D cartoon' aesthetic much better and followed all text and icon instructions perfectly, making it the more successful image for this specific 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.”
AI Judge Analysis
FLUX.1 [dev]
- + Warm, golden lighting creates a very cohesive atmosphere.
- + Beautiful bokeh effect with well-rendered butterflies.
- − Failed the prompt by missing the tabby kitten entirely.
- − Features an 'animated' or 'Disney' style instead of the requested hyper-photorealism.
- − The animals look like stylized 3D characters rather than real babies.
Qwen Image
- + Successfully included all four requested animals: golden puppy, tabby kitten, bunny, and fox kit.
- + Achieved much better photorealism with natural fur textures and lighting.
- + Excellent inclusion of requested details like god rays and dew sparkles on the grass.
- − One butterfly is quite large relative to the animals.
- − The large dog in the center is slightly less 'puppy-like' compared to the others.
Verdict: Qwen Image is the clear winner as it strictly adhered to the prompt by including all four specific animals, whereas FLUX.1 [dev] completely omitted the kitten. Furthermore, Qwen Image delivered a photorealistic result as requested, while FLUX.1 [dev] produced a heavily stylized, cartoonish aesthetic.
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.1 [dev]
- + Excellent typography style that fits the vintage aesthetic.
- + Clean vector emblem layout with elegant banner detailing.
- + Includes all requested elements like the cloche and 'Est. 1720' with professional balance.
- − Misspelled words including 'Flariláan' instead of 'Florian' and 'RESEAURANT'.
- − Added unnecessary numbers ('11011', '1941') that clutter the design.
Qwen Image
- + Correct spelling of 'Caffé' and clearly legible 'Est. 1720' text.
- + Stronger adherence to the 'minimalist' instruction with a bold, simple cloche icon.
- + Good use of subtle paper texture on the background as requested.
- − The main name 'Florian' is awkwardly broken up and illegible in the middle.
- − The composition feels slightly bottom-heavy compared to the top.
Verdict: FLUX.1 [dev] produces a much more professional-looking vector emblem that captures the vintage restaurant aesthetic perfectly, but suffers from significant spelling errors and hallucinations of extra numbers. Qwen Image manages the spelling of the smaller text much better and captures the requested texture, but the graphical arrangement of the main brand name is messy and difficult to read. FLUX.1 [dev] is the likely winner as its design is more salvageable for a logo context.
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.1 [dev]
- + Elegant layout with a professional, symmetrical design.
- + Captures the sophisticated 'NASA-inspired' aesthetic with muted colors and crisp vector lines.
- + Consistent iconography at the bottom of the poster.
- − Text is largely illegible gibberish despite the correct visual style.
- − Includes extra planets like Saturn which are not relevant to the Apollo 11 mission.
Qwen Image
- + Legible text that correctly identifies crew members and mission milestones.
- + Clearer sequence of steps with numbered labels.
- + Includes a landing site marker for 'Tranquility' as requested by the context of a mission infographic.
- − Includes literal prompt instructions (Stop at landing) in the final image text.
- − The layout is more cluttered and less 'modern' compared to Model A.
- − Contains some typos like 'Sarth' and 'Apoll.'
Verdict: FLUX.1 [dev] much more closely adheres to the requested 'modern vector' style and NASA color palette, creating a visually pleasing poster that looks professional despite the nonsensical text. Qwen Image performs better at functional communication with legible names and clear steps, but its design is more juvenile and it failed to filter out meta-instructions from the prompt text. FLUX.1 [dev] is the preferred choice for a high-quality infographic base.
Explore each model
Alibaba's Qwen image model