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
22.9
arena score
#30 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Stable Diffusion 3.5 Large Turbo
10.7
arena score
#61 of 62 in Text-to-Image
Vote tally
Where the votes landed
Qwen Image 2512
0%
win rate
Ties
0%
Stable Diffusion 3.5 Large Turbo
0%
win rate
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Perfect adherence to object placement instructions
- + Highly realistic textures for the wood, book, and glass
- + Excellent handling of reflections and refractions through the glass cube
- − The glass cube has an unusually thick, slightly teal-tinted base
Stable Diffusion 3.5 Large Turbo
- + Clean, modern aesthetic with sharp lines
- + Accurate lighting direction from the left
- − Failed the spatial layout by putting the book inside the cube instead of on top
- − The plant is beside the cube rather than behind it as requested
- − Render quality looks more like a digital CGI model than a photograph
Verdict: Qwen Image 2512 followed every spatial instruction perfectly, placing the sphere inside the cube and the book on top, while maintaining a high level of photographic realism. Stable Diffusion 3.5 Large Turbo failed on the core spatial prompt, placing the book inside the cube and the sphere on top of the book, and the overall image quality is significantly less realistic.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent photorealistic skin texture and facial details
- + Accurate depiction of a 50mm lens with shallow depth of field and beautiful bokeh
- + Perfect adherence to the 'cinematic but realistic' and 'no stylization' constraints
- − The subject is posing/looking at the camera, which slightly conflicts with the 'candid' prompt
- − Minor anatomical inconsistencies with how the hands are resting on the bike
Stable Diffusion 3.5 Large Turbo
- + Captures the action of repairing better than Model A
- + Good depiction of the red bicycle
- − Very poor skin texture and facial rendering, appearing plastic-like
- − Fails to provide the requested bokeh/shallow depth of field; cars in background are too sharp
- − The hat and hair texture are highly artificial and low quality
Verdict: Qwen Image 2512 produces a stunningly realistic photograph that captures the specific lighting, lens effects, and skin textures requested in the prompt. Stable Diffusion 3.5 Large Turbo fails to meet the quality requirements, producing an image that looks like a low-fidelity CGI render with flat textures and no realistic depth of field.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent adherence to all prompt details including beads, scars, and leather textures.
- + Superior lighting effects with realistic torchlight reflections and bokeh sparks.
- + High cinematic quality with lifelike eyes and skin texture.
- − One braid on the right side of the face appears to merge awkwardly with the armor.
Stable Diffusion 3.5 Large Turbo
- + Clear engraved patterns on the armor.
- + Highly stylized and clean braids.
- − Missed the 'beads' in the hair and the 'leather straps' requested in the prompt.
- − The 'scars and dirt' look more like fresh blood or paint rather than weathered grit.
- − Plastic-like skin texture and flat lighting compared to the requested lilelike quality.
Verdict: Qwen Image 2512 is the clear winner as it captured every nuance of the prompt, including the fine details of the leather straps, beads, and the specific atmosphere of warm torchlight. Stable Diffusion 3.5 Large Turbo produced an image that feels more like a 3D render with artificial-looking skin and lighting, failing to incorporate several of the smaller textural requests.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent structure that resembles a real menu layout.
- + High-quality, realistic food photography integrated into the grid.
- + Good use of bold typography and vibrant accents as requested.
- − Text is mostly gibberish despite looking like a font.
- − Slight spelling errors in headers like 'Appetiizers' and 'Means'.
Stable Diffusion 3.5 Large Turbo
- + Features a very clean, high-contrast white background.
- + Visuals are vibrant and colorful.
- + Text is more legible, though still contains a typo in 'Mians'.
- − The layout is cluttered and does not feel like a single cohesive menu design.
- − Food images look more like AI-generated abstract art than realistic restaurant photos.
- − Composition is fragmented with items floating and cropped poorly.
Verdict: Qwen Image 2512 is the clear winner as it successfully designs a cohesive, professional restaurant menu that follows all layout instructions. While Stable Diffusion 3.5 Large Turbo provides a clean aesthetic, it fails to produce a functional-looking document, resulting in a disorganized collection of images rather than a menu.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent typography with perfect rendering of all three requested text fields.
- + Highly photorealistic textures on the meat patty, vegetables, and bun.
- + Dynamic composition with effective 'exploded' effect and flying debris.
- − The 'starburst' for the price tag is more of a jagged badge but still fits the theme.
Stable Diffusion 3.5 Large Turbo
- + Strong vibrant colors and a dramatic fire effect.
- − Completely failed to include the requested text elements.
- − The burger is not 'exploded' as requested, but rather stacked with a stick through it.
- − Less photorealistic and more of a digital illustration style.
Verdict: Qwen Image 2512 followed every instruction perfectly, including complex text rendering and a difficult exploded composition. Stable Diffusion 3.5 Large Turbo failed to include any of the text and did not adhere to the 'exploded' layout, resulting in a generic burger image.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent text rendering with accurate spelling of complex menu items.
- + Authentic chalkboard texture including smudges and realistic chalk dust.
- + Consistently elegant cursive style that matches the prompt's aesthetic.
- − Minor spelling error in 'Risitto' instead of 'Risotto'.
- − The text looks slightly like a digital font overlay due to perfectly clean edges on some letters.
Stable Diffusion 3.5 Large Turbo
- + Bright, clean composition with good lighting effects.
- + Includes the requested cafe environment elements like plants and lighting.
- − Significant spelling failures and garbled text across the entire board.
- − Text style appears more like a digital marker or vector than realistic chalk.
- − Failed to include the specific dates and prices requested in the prompt.
Verdict: Qwen Image 2512 is the clear winner as it successfully rendered almost all of the complex text requested with high legibility and a realistic chalkboard aesthetic. In contrast, Stable Diffusion 3.5 Large Turbo struggled with basic spelling and failed to follow the specific layout and content instructions of the prompt.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent photorealism in the textures of the horse's coat and the astronaut's suit.
- + Superior cinematic lighting and focus depth.
- + High level of detail in the astronaut's visor and equipment.
- − Failed the negative constraint; the astronaut is riding the horse, not vice versa.
- − The horse's hind legs are oddly positioned, merging into the bottom of the frame.
Stable Diffusion 3.5 Large Turbo
- + Stronger adherence to the surreal aspect with stylized clouds and composition.
- + Clean, high-contrast visual style.
- − Failed the negative constraint; the astronaut is riding the horse, not vice versa.
- − The horse's anatomy is distorted, particularly the elongated neck and disjointed rear leg.
- − The Earth in the background is poorly rendered and lacks detail.
Verdict: Both Qwen Image 2512 and Stable Diffusion 3.5 Large Turbo failed to follow the specific 'horse on top' instruction, providing standard astronaut-on-horse images. Qwen Image 2512 is the winner because it provides significantly better photographic quality, realistic textures, and cinematic detail compared to the anatomically awkward and flatter rendering from Stable Diffusion.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent photorealism in both textures and lighting
- + Captures the woman's bored expression perfectly as requested
- + Detailed capybara fur and professional driver uniform
- − The paws on the steering wheel look more like primate hands than capybara paws
- − The camera angle is a bit cramped, making the taxi look small
Stable Diffusion 3.5 Large Turbo
- + Natural composition with a better sense of depth and car interior space
- + Correct depiction of capybara-like paws on the wheel
- + Dynamic city bokeh effect through the window
- − Failed to include the woman looking at her phone; she is just looking ahead
- − The woman is out of focus and much less detailed than in the other image
- − Text on the cap is slightly distorted
Verdict: Qwen Image 2512 followed all instructions perfectly, including the specific requirement for the passenger to be on her phone with a bored expression. While Stable Diffusion 3.5 Large Turbo has a more cinematic side-profile composition and more accurate animal paws, it missed a core narrative element of the prompt regarding the human character's action.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent adherence to the text prompt requirements, including dates and location.
- + High-quality cinematic lighting and atmosphere.
- + Good execution of the requested 'thorns and webs' border and scroll banner.
- − One minor spelling error in the word 'Hallowern'.
Stable Diffusion 3.5 Large Turbo
- + Successfully captures the parchment paper texture.
- + Clean, graphic-style illustration.
- − Failed to include most of the requested text, including the invite message, date, time, and location.
- − The title text 'Halloween Party' is incomplete and uses an unusual font for the final letter.
- − Composition feels a bit empty and lacks the requested gothic 'moody' sky.
Verdict: Qwen Image 2512 is the clear winner as it followed almost all instructions, including complex text details and specific design elements like the scroll banner and thorn border. Stable Diffusion 3.5 Large Turbo failed to include the event details and the scroll, resulting in a much simpler and less functional image.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent text accuracy for both 'JAPAN' and 'SUSHI'.
- + High-quality textures with convincing 'cartoon-realistic' PBR materials.
- + Follows the aesthetic of a miniature diorama perfectly with cohesive lighting and composition.
- − The flag icon is a separate element next to the text rather than integrated smoothly.
Stable Diffusion 3.5 Large Turbo
- + Strong isometric perspective and clean 3D modeling style.
- + Creative use of signage for the text elements.
- − Spelling error in the text, rendering 'SUSHI' as 'SIIHI'.
- − The flag icon is incorrect, resembling a generic red/white horizontal split rather than the Japanese flag.
- − Background is a dull grey rather than the requested solid light blue.
Verdict: Qwen Image 2512 followed every instruction in the prompt, including the specific light blue background and accurate text rendering. Stable Diffusion 3.5 Large Turbo failed on the background color, the spelling of 'SUSHI', and did not provide an accurate Japanese 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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Successfully included all four requested animals (dog, cat, rabbit, fox).
- + High level of photographic realism in fur texture and lighting.
- + Beautiful inclusion of god rays and dew sparkles as requested.
- − The fox's anatomy is slightly merged with the dog's rear.
- − The animals are posed more for a static portrait than 'chasing or tumbling'.
Stable Diffusion 3.5 Large Turbo
- + Bright, vibrant colors that fit a 'wholesome' vibe.
- + Good rim lighting effect on the animals' fur.
- − Failed to include the bunny and fox kit entirely.
- − The aesthetic is more illustrative/cartoonish than hyper-photorealistic.
- − Anatomical issues with the paws of the animal on the left.
Verdict: Qwen Image 2512 followed the prompt almost perfectly, including all four specific animals with high-quality, realistic textures and the requested lighting effects. Stable Diffusion 3.5 Large Turbo failed the prompt adherence significantly by missing two of the four animals and adopting an illustrative style rather than the requested photorealism.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Excellent text rendering with no spelling errors
- + Beautiful engraving style with high illustrative quality
- + Strong adherence to all prompt elements including steam and the banner
- − The steam is a bit over-elaborate for a 'minimalist' logo request
Stable Diffusion 3.5 Large Turbo
- + Closer to a minimalist vector style
- + Good use of negative space and classic color tones
- − Spelling error in the main text ('Caffeé')
- − The cloche is awkward and looks combined with a coffee mug
- − Typography is less cohesive and has artifacts on the 'n'
Verdict: Qwen Image 2512 followed the prompt much more effectively, delivering perfect text and a well-composed vintage emblem. Stable Diffusion 3.5 Large Turbo struggled with the requested name and produced a confusing hybrid object that was neither a clear cloche nor a mug.
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.”
Qwen Image 2512
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
Qwen Image 2512
- + Successfully included detailed illustrations for the Saturn V and Lunar Module.
- + Adhered well to the sequential 6-step layout requested in the prompt.
- + The 'landing' section includes the names of the three astronauts (Armstrong, Aldrin, Collins).
- − Several spelling errors in the labels (e.g., 'Translaurtcit', 'Desceeint').
- − Instruction text 'Steps stop at landing' was accidentally included as a title element.
Stable Diffusion 3.5 Large Turbo
- + Strong aesthetic appeal with a consistent retro-vector illustration style.
- + Excellent adherence to the color palette and NASA-inspired design language.
- + Clean layout with sophisticated graphic design elements.
- − Failed to provide the requested 6-step sequential content.
- − Text is largely 'lorem ipsum' style gibberish or heavily misspelled (e.g., 'Apoll.o', 'Laurch').
- − Missing specific icons like the Saturn V requested in the prompt.
Verdict: Qwen Image 2512 is the more functional infographic, as it successfully attempts to follow the 6-step sequence and includes specific technical illustrations requested, despite several typos. Stable Diffusion 3.5 Large Turbo produces a more visually striking piece of graphic art that captures the 'vibe' and palette perfectly, but it fails to include the requested instructional steps and produces nonsensical text throughout.
Next steps
Explore each model
Distilled version of SD 3.5 Large that generates high-quality images in just 4 steps, offering faster inference and reduced costs