Improved version of Alibaba's Qwen image model with better text rendering, finer natural textures, and more realistic human generation.
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
Qwen Image 2512
#30 of 62 in Text-to-Image
Stable Diffusion 3.5 Large
#29 of 62 in Text-to-Image
Where the votes landed
Qwen Image 2512
80.0%
win rate
Ties
0.0%
Stable Diffusion 3.5 Large
20.0%
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
Qwen Image 2512
- + Perfectly follows the spatial instruction of the book sitting on top of the cube.
- + Excellent rendering of refraction and reflections within the glass panels.
- + Very realistic lighting and natural depth of field.
- − The glass has a slightly green tint rather than being perfectly clear.
Stable Diffusion 3.5 Large
- + Very clean, high-resolution rendering of the blue sphere.
- + Strong adherence to the 'soft window light' lighting instruction.
- − Failed the spatial instruction: the book is inside the cube instead of on top.
- − The placement of the sphere on top of the book contradicts 'inside the cube' if the book is considered the base.
- − The glass cube lacks a bottom edge, appearing to just sit around the book.
Verdict: Qwen Image 2512 followed all spatial instructions perfectly, accurately placing the red book on top of the glass cube and the sphere inside. Stable Diffusion 3.5 Large failed the core prompt logic by placing the book inside the cube and the sphere on top of the book, which also resulted in a less convincing physical setup.
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
Qwen Image 2512
- + Exceptional realism in skin texture and facial details.
- + Perfectly captures the 'candid street photo' look with natural lighting.
- + Successfully incorporates motion blur from passing cars in the background.
- − The man is posing/looking at the camera rather than actively repairing the bike.
- − The bike is somewhat simplified in its mechanical structure.
Stable Diffusion 3.5 Large
- + Shows the subject actively engaged in the task of repairing/working on the bike.
- + Good atmospheric rain effects and reflections on the pavement.
- + Captures the full silhouette of the bike and the man.
- − The man's proportions and arms look slightly distorted/wiry.
- − The background car lacks the requested motion blur, appearing more static.
- − The image has a more 'digital' feel compared to the requested 50mm lens look.
Verdict: Qwen Image 2512 wins due to its incredible photographic realism and adherence to the 'natural skin texture' and '50mm lens' prompts, looking like a genuine street photograph. While Stable Diffusion 3.5 Large better captured the action of 'repairing', it suffered from slight anatomical irregularities and a less convincing cinematic texture.
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
Qwen Image 2512
- + Excellent adherence to the 'beads in hair' prompt element.
- + Strong, lifelike skin texture with realistic scarring and dirt.
- + Masterful use of lighting and bokeh to create a cinematic atmosphere.
- − The leather straps slightly overlap the armor in a way that looks a bit cluttered.
Stable Diffusion 3.5 Large
- + Ornate engraving on the armor is very intricate and well-defined.
- + Good textures on the gambeson and chainmail elements.
- + Clear adherence to the battle-worn aspect with dirt and blood.
- − Missed the 'small beads' in the hair completely.
- − The lighting feels a bit secondary compared to the background elements.
- − Face looks slightly airbrushed compared to the rugged skin in Model A.
Verdict: Qwen Image 2512 followed every detail of the prompt, including specific details like the beads in the braids, and produced a more lifelike portrait with superior lighting. Stable Diffusion 3.5 Large has impressive armor detail but failed to include the requested beads and the overall image quality feels slightly more synthetic.
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
Qwen Image 2512
- + Features a clear grid layout of food photos as requested.
- + Organized categorical sections (Appetizers, Mains, etc.) with pricing.
- + Good use of bold sans-serif fonts and vibrant accent colors.
- − Text contains several spelling errors (e.g., 'APPEITIIZERS', 'MEANS').
- − The small body text is somewhat garbled and contains artifacts.
Stable Diffusion 3.5 Large
- + High-quality food photography with a professional, clean aesthetic.
- + Excellent implementation of bold sans-serif typography.
- + Minimalist design creates a premium look for a casual dining setting.
- − Layout is more of a spread than a single-page menu design.
- − Text significantly degrades into gibberish in the lower sections.
- − Food photos are cropped at the edges of the frame.
Verdict: Qwen Image 2512 produces a much more functional menu layout that follows the 'sections' part of the prompt more logically. While Stable Diffusion 3.5 Large has better individual photo quality and a sleek aesthetic, it feels more like a magazine spread than a usable restaurant menu, whereas Qwen Image 2512 captures the specific grid-and-list structure requested.
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
Qwen Image 2512
- + Perfect text rendering according to the prompt instructions.
- + Dynamic 'exploded' view accurately shows suspended components.
- + Professional commercial layout with a clear hierarchy of information.
- − The patty texture is a bit uniform compared to the other ingredients.
Stable Diffusion 3.5 Large
- + High level of photorealism in the food textures and melting cheese.
- + Great lighting interaction between the flames and the burger bun.
- − Completely failed to include any of the requested text.
- − Did not follow the 'exploded' instruction; the burger is assembled and floating.
- − Lack of 'starburst' element for the pricing.
Verdict: Qwen Image 2512 followed every instruction in the prompt, including complex text rendering with specific effects and the 'exploded' layout. Stable Diffusion 3.5 Large produced a high-quality food image but failed to include any text or the exploded component structure, making it unsuitable for the requested advertisement design.
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
Qwen Image 2512
- + Excellent text legibility and spelling accuracy for almost all specified items.
- + Beautiful cursive handwriting style that perfectly matches the 'chalkboard' aesthetic.
- + Realistic chalk texture including smudges and dust on the board.
- − Minor spelling error in 'Risotto' (spelled as Risitto).
- − The composition is a very tight close-up, sacrificing some of the 'cozy café' environmental context.
Stable Diffusion 3.5 Large
- + Includes more of the 'cozy café' environment with furniture and plants.
- + Good chalky texture on the board and surrounding wall.
- − Significant spelling errors throughout the text (e.g., 'Todaay', 'Ottpups', 'Browal').
- − Failed to follow the specific item requests, mixing words into different sections.
- − Does not use the requested cursive handwriting style for the body text.
Verdict: Qwen Image 2512 is the clear winner as it successfully rendered nearly all requested text with high legibility and a beautiful, consistent cursive chalk style. In contrast, Stable Diffusion 3.5 Large struggled significantly with text coherence, producing multiple misspellings and failing to maintain the requested elegant handwriting format.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
Qwen Image 2512
- + Excellent high-resolution detail on the spacesuit and horse's coat.
- + Clear, realistic facial features visible through the visor.
- + Clean composition with sharp focus on the subjects against the curved horizon of Earth.
- − Failed the specific positional prompt instruction; the astronaut is riding the horse instead of the horse riding the astronaut.
- − Anatomy of the horse's rear leg is slightly awkward/truncated.
Stable Diffusion 3.5 Large
- + Beautiful cinematic lighting and atmosphere with stardust effects.
- + Impressive sense of scale and movement.
- + Accurate rendering of space elements like nebulae and city lights from orbit.
- − Failed the specific positional prompt instruction; the astronaut is riding the horse.
- − The horse's muzzle has some minor structural artifacts.
Verdict: Both models failed the complex 'surreal' logic of the prompt, which specifically requested the horse be on top of the astronaut. Between the two standard interpretations, Qwen Image 2512 provides much higher clarity and realistic skin/material textures, while Stable Diffusion 3.5 Large offers a more artistic and atmospheric composition.
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
Qwen Image 2512
- + Excellent adherence to the scene composition, showing both the driver and the passenger as requested.
- + Realistic capybara fur texture and professional expression.
- + Accurately depicts the bored expression of the businesswoman in the rear.
- − The paws on the steering wheel look slightly anthropomorphized (like human hands in gloves).
Stable Diffusion 3.5 Large
- + High-quality lighting and vibrant colors in the taxi interior.
- + Effective use of bokeh for the city lights in the background.
- − Failed to include the human businesswoman in the back seat entirely.
- − The capybara's anatomy is distorted with human-style legs and jeans.
- − The capybara's front 'paws' are not positioned correctly on the steering wheel.
Verdict: Qwen Image 2512 followed the prompt instructions much more effectively by including the bored businesswoman in the backseat, which was a key narrative element. In contrast, Stable Diffusion 3.5 Large missed the passenger entirely and created a confusing anatomical hybrid for the capybara, adding human legs and jeans that were not requested.
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
Qwen Image 2512
- + Excellent typographic arrangement and legible event details
- + Highly detailed thorns and spiderweb border following the prompt exactly
- + Cinematic lighting on the central jack-o-lantern
- − Spelling error in the main title ('Hallowern' instead of 'Halloween')
- − The background is more of a digital illustration than a 'parchment poster' texture
Stable Diffusion 3.5 Large
- + Strong 'dark parchment' aesthetic with torn edges and vintage texture
- + Creative use of a large moon to create high-contrast silhouettes
- + Captures the gothic aesthetic well with graveyard-style elements
- − Failed to include the requested event details (Date, Time, Location)
- − Garbled and illegible text at the bottom of the scroll
- − Composition is a bit cluttered with multiple small pumpkins instead of one central focus
Verdict: Qwen Image 2512 followed the layout and functional requirements of the invitation much better, including all requested event details clearly, though it suffered from a typo in the main header. Stable Diffusion 3.5 Large succeeded in capturing the parchment texture and vintage vibe, but failed significantly on the text-heavy requirements of the prompt. Qwen Image 2512 is the winner for providing a usable, albeit slightly misspelled, invitation 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
Qwen Image 2512
- + Excellent text rendering with 'JAPAN' and 'SUSHI' exactly as requested.
- + Perfectly captures the 'miniature 3D cartoon' aesthetic with smooth, clean textures.
- + Highly professional composition mirroring a polished mobile game or 3D icon.
Stable Diffusion 3.5 Large
- + Good isometric perspective and diorama base structure.
- + Vibrant colors and a high variety of sushi types.
- − Failed the text placement requested, attaching it to a flag instead of top-center.
- − Texture quality feels more like plastic than 'refined PBR' materials.
- − The text on the main flag is slightly aliased and low-quality compared to the rest of the scene.
Verdict: Qwen Image 2512 followed the prompt instructions near-perfectly, placing the text and flag icon exactly where requested with a very clean, high-quality 3D render style. Stable Diffusion 3.5 Large struggled with the text placement and rendered the scene with a busier, less refined look that missed the 'minimal garnish' requirement.
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
Qwen Image 2512
- + Excellent fur texture and fine detail on all four animals.
- + Successfully includes all four requested animals (dog, cat, bunny, fox) in a clear composition.
- + Very high resolution with sharp focus on the central subjects.
- − The composition is a static 'family portrait' style rather than the requested 'playfully chasing' and 'tumbling' action.
- − The lighting feels a bit more artificial/digital compared to the more natural atmosphere in the other image.
Stable Diffusion 3.5 Large
- + Captures the 'chasing' and 'tumbling' movement requested in the prompt perfectly.
- + Breathtaking atmospheric lighting with beautiful bokeh, dew sparkles, and god rays.
- + High level of cuteness and dynamic energy that fits the 'joyful wholesome vibe'.
- − The kitten's anatomy is slightly generic and looks more like a small fox/cat hybrid.
- − Slightly lower sharpness on the fur textures compared to the other model.
Verdict: While Qwen Image 2512 produces a sharper image with better-defined textures for each specific animal, it fails to capture the 'chasing and tumbling' action requested. Stable Diffusion 3.5 Large is the superior choice here because it perfectly interprets the dynamic energy and atmospheric lighting of the prompt, creating a much more evocative and 'wholesome' scene.
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
Qwen Image 2512
- + Excellent typography with correct spelling of 'Caffè'.
- + Highly detailed and polished vector emblem style with great use of shading.
- + Perfect integration of the 'Est. 1720' banner into the overall composition.
- − The detail level leans more toward'vintage illustration' than 'minimalist' logo.
Stable Diffusion 3.5 Large
- + Closer adherence to the 'minimalist' aspect of the prompt with flat icons.
- + Subtle texture on the background adds a nice aged paper feel.
- − Spelling error in 'Cafféé Florian' with an extra 'e'.
- − The steam and cloche icons are poorly integrated and visually disjointed.
- − The typography on the banner is less elegant than Model A.
Verdict: Qwen Image 2512 produces a much more professional and aesthetically pleasing logo with perfect spelling and high-quality vintage typography. While Stable Diffusion 3.5 Large attempts a more minimalist approach, it fails due to a spelling error ('Cafféé') and a clunky graphic design where elements appear floating and disconnected.
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
Qwen Image 2512
- + Excellent adherence to the logical steps requested in the prompt.
- + Clean, modern vector illustration style with high visual appeal.
- + Mostly legible text including names and correctly numbered steps.
- − Several spelling errors in the smaller labels (e.g., 'TranslauraJ', 'Desceeint').
- − Duplicated the '2. Earth Orbit' label.
Stable Diffusion 3.5 Large
- + Strong 'NASA-inspired' aesthetic and professional layout design.
- + Good use of the requested color palette.
- − Included a Space Shuttle instead of the requested Saturn V rocket.
- − Garbled, unreadable text throughout the entire infographic.
- − Failed to follow the requested numbered steps for the mission.
Verdict: Qwen Image 2512 is the clear winner as it followed the technical instructions for the mission steps and included the correct hardware (Saturn V and Lunar Module), whereas Stable Diffusion 3.5 Large produced a generic space-themed poster with a Space Shuttle and illegible text. While Qwen has some spelling issues, its composition actually functions as the requested infographic.
Explore each model
Stability AI's 8.1-billion parameter Multimodal Diffusion Transformer (MMDiT) text-to-image model featuring improved image quality, typography, complex prompt understanding, and resource-efficiency