OpenAI's legacy image generation model supporting generations, edits with masks (inpainting), and variations
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
DALL-E 2
15.8
arena score
#59 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Stable Diffusion 3.5 Large
22.9
arena score
#29 of 62 in Text-to-Image
Vote tally
Where the votes landed
DALL-E 2
0.0%
win rate
Ties
0.0%
Stable Diffusion 3.5 Large
100.0%
win rate
0.0%
0.0% ties
100.0%
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Features a wooden table surface with realistic reflections.
- + Correct soft window lighting.
- − Failed almost all spatial instructions, including the blue sphere and red book placement.
- − The plant is in a blue pot rather than behind the cube.
- − The cube appears solid with a red center rather than empty with a sphere inside.
Stable Diffusion 3.5 Large
- + Excellent adherence to spatial prompts, placing all items in the correct relative positions.
- + High visual clarity and realistic glass refractions.
- + Accurately depicts the plant behind the cube visible through the glass.
- − The sphere appears to be inside the cube but floating/sitting on the book, which was positioned 'on top' of the cube in the prompt.
- − The 'red book' is inside the cube rather than on top of it.
Verdict: Stable Diffusion 3.5 Large followed the complex spatial instructions much more effectively than DALL-E 2, which failed to render the blue sphere or the red book correctly. While Stable Diffusion 3.5 Large swapped the vertical order of the book and the cube's interior, it successfully captured all requested elements and lighting effects in a high-fidelity image.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Captured a unique, voyeuristic 'imperfect framing' angle
- + Effective wet pavement reflections
- − Extremely poor resolution and clarity
- − Failed to show the subject's face or ethnicity as requested
- − Depth of field is so shallow it obscures almost all relevant detail
Stable Diffusion 3.5 Large
- + Excellent adherence to all prompt elements including age, ethnicity, and garment texture
- + Beautiful rendering of rain, wet surfaces, and bokeh
- + High visual quality with realistic skin and environmental details
- − Missed the 'motion blur from passing cars' request as vehicles appear static
- − The background bus has some minor structural incoherence
Verdict: Stable Diffusion 3.5 Large successfully interprets nearly all aspects of the complex prompt, delivering a cinematic and high-resolution image with realistic textures and lighting. In contrast, DALL-E 2 produced a low-quality, blurry output that failed to clearly depict the primary subject or many of the requested details. While Stable Diffusion 3.5 Large missed the motion blur for the background traffic, its overall composition and technical execution are far superior.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Strong bokeh effect
- + Gritty, battle-worn texture
- − Lack of clear facial features and eyes
- − Anatomically confusing structures
Stable Diffusion 3.5 Large
- + Highly detailed engraved armor and textures
- + Excellent adherence to braided hair and scars
- + Clear, lifelike eyes and facial structure
- − Less of a 'close' portrait than requested
- − Braids lack the specific beads mentioned
Verdict: Stable Diffusion 3.5 Large successfully captures nearly all prompt elements including fine armor engravings, lifelike eyes, and subtle dirt. DALL-E 2 produced a highly abstract and distorted image that failed to render a recognizable paladin or the requested details clearly. Stable Diffusion 3.5 Large is the clear preference for its technical execution and adherence.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
0% wins
0% ties
100% wins
AI Judge Analysis
DALL-E 2
- + Strong bold sans-serif typography on the right side.
- + Good color vibrancy in the food-inspired mosaics.
- − The food photos are fragmented into abstract shards rather than a clear menu grid.
- − The text is completely nonsensical and gibberish.
- − Layout does not resemble a functional restaurant menu.
Stable Diffusion 3.5 Large
- + Excellent adherence to the menu layout with clear sections for Appetizers and Mains.
- + High-quality, distinct food photos arranged in a professional grid.
- + Text is largely readable and follows the bold sans-serif request.
- − Slight spelling artifacts in category headers like 'Appetizrs'.
- − The grid layout feels slightly repetitive with multiple similar pizza images.
Verdict: Stable Diffusion 3.5 Large far outperforms DALL-E 2 by providing a functional, professional menu design that accurately interprets all aspects of the prompt, including specific food sections and a grid layout. DALL-E 2 produced an abstract, messy composition that lacks the clarity and structural requirements of a restaurant 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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Successfully integrated a glowing title text.
- + Conveys a sense of explosion and motion with deconstructed layers.
- − Text is misspelled ('MARGIC BAGUEC').
- − Image quality is blurry and lacks photorealistic detail.
- − Failed to include the price and secondary messages requested.
Stable Diffusion 3.5 Large
- + Excellent photographic quality with realistic textures on the patty and vegetables.
- + Captures the fiery background and glowing embers perfectly.
- + The lighting and composition are professional and visually appealing.
- − Failed to include any of the requested text.
- − The burger is stacked rather than 'exploded' with components suspended in mid-air.
Verdict: DALL-E 2 attempted the complex text requirements but failed significantly on image quality and spelling. Stable Diffusion 3.5 Large produced a stunning, high-quality image that correctly captured the fiery aesthetic, but it completely ignored the text prompts and the 'exploded' layout. Stable Diffusion 3.5 Large is the preferred choice for its superior visual fidelity and professional execution of the atmosphere.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Successfully captures a realistic chalk-on-board texture.
- + Displays handwritten variation in letter sizes and slants.
- − Fails completely to render legible English text or follow the specific menu prompt.
- − The image is low resolution and poorly cropped.
- − Does not show the café environment requested.
Stable Diffusion 3.5 Large
- + Excellent environmental context with a realistic and cozy café interior.
- + High resolution with clean, professional composition.
- + Follows the prompt's layout and captures the correct date and several menu keywords.
- − Text contains several spelling errors (e.g., 'Todaay', 'Cholcalte').
- − Handwriting looks slightly more like a digital font than natural chalk for certain headings.
Verdict: DALL-E 2 fails significantly on this task, producing unintelligible gibberish and a low-quality crop that ignores most of the prompt. Stable Diffusion 3.5 Large provides a high-quality, professional image of a café that mostly adheres to the text requirements, despite some minor spelling mistakes and a slightly more rigid handwriting style.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Successfully captures a dreamy, surreal aesthetic.
- + Composition is simple and balanced.
- − Failed the core spatial logic of the prompt ('horse on top').
- − Low resolution with significant noise and artifacts.
- − Anatomy of the horse and astronaut is poorly defined.
Stable Diffusion 3.5 Large
- + High visual quality with cinematic lighting and detail.
- + Excellent rendering of textures like the space suit and clouds.
- − Failed the core spatial logic of the prompt ('horse on top').
- − The horse has a distorted leg near the clouds.
- − The image is a generic interpretation of 'astronaut on a horse' despite the specific directive.
Verdict: Both models failed the negative/spatial constraint of placing the horse on top of the astronaut, instead defaulting to the common trope of an astronaut riding a horse. Stable Diffusion 3.5 Large is the clear winner solely on visual merit, offering cinematic quality and intricate details that DALL-E 2 cannot match.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Attempts to include both the driver and the passenger as requested
- − Severe anatomical distortions in the human and animal figures
- − Heavy visual artifacts and blurry textures
- − Fails to realistically depict a capybara, appearing more like an abstract sculpture
- − Perspective and lighting are incoherent
Stable Diffusion 3.5 Large
- + High image quality with realistic textures and fur detail
- + Perfectly executes the capybara's outfit, including the taxi cap and jacket
- + Consistent lighting and professional composition
- − Completely misses the second half of the prompt regarding the businesswoman in the back seat
- − Only shows one paw on the steering wheel instead of two
Verdict: DALL-E 2 followed more of the prompt by including the passenger, but produced an image with such low visual quality and extreme distortion that it is unusable. Stable Diffusion 3.5 Large produced a high-quality, photorealistic image of the capybara but failed to render the interior passenger and back seat, focusing only on the driver.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Strong vintage aesthetic
- + Atmospheric, moody lighting
- − Text is largely illegible and contains numerous misspellings
- − Missing nearly all specific prompt details like the jack-o-lantern and event details
- − Poor image clarity and resolution
Stable Diffusion 3.5 Large
- + Excellent adherence to all prompt elements including webs, thorns, and jack-o-lanterns
- + Accurate rendering of the title and scroll banner text
- + High visual clarity and polished composition
- − Missing the specific event details (Date, Time, Location) at the bottom
- − Text at the very bottom of the scroll is garbled
Verdict: Stable Diffusion 3.5 Large far outperforms DALL-E 2 by providing a clear, high-resolution image that follows almost all prompt instructions. While DALL-E 2 captures a moody atmosphere, its complete failure to render legible text or specific objects like the jack-o-lantern makes it unusable for an invitation, whereas Stable Diffusion 3.5 Large produces a professional-looking graphic.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Clean isometric perspective
- + Vibrant colors matching the background request
- − Incomplete and misspelled text ('Sush') placed on the plate instead of top-center
- − Failed to include 'JAPAN' text or a flag icon
- − The sushi models are extremely abstract and lack detail
Stable Diffusion 3.5 Large
- + Perfect adherence to text requirements ('JAPAN', 'SUSHI', and flag icon)
- + Excellent 3D miniature aesthetic with high-quality PBR textures for the rice and fish
- + Correct isometric composition on a diorama base
- − Slightly more garnish than requested in the 'minimal' instruction
- − Text is on a sign rather than floating at the top-center of the image
Verdict: Stable Diffusion 3.5 Large followed every part of the prompt, including complex text rendering and specific scene elements like the diorama base. DALL-E 2 failed significantly on the text, spelling 'Sush' and omitting 'JAPAN' entirely, while also producing very low-detail models. Stable Diffusion 3.5 Large is the clear winner for its superior visual quality and prompt adherence.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Natural look to the golden retriever fur texture.
- + Strong sense of dynamic motion with the animal's pose.
- − Severely malformed anatomy on the cat and fox in the background.
- − Major artifacts in the butterfly and grass rendering.
- − Lacks the requested bunny and consistent lighting quality.
Stable Diffusion 3.5 Large
- + Includes all four requested animals with clear, distinct features.
- + Excellent lighting including god rays, bokeh, and dew sparkles.
- + High visual coherence with expressive eyes and soft fur textures.
- − Central puppy's paw is slightly blurred/blended into the grass.
- − The fox and cat have very similar facial structures.
Verdict: Stable Diffusion 3.5 Large successfully captures all four specified animals with high detail and beautiful atmospheric lighting that matches the 'wholesome' prompt perfectly. DALL-E 2 fails significantly on technical quality, producing nightmare-like anatomical distortions on the background animals and missing the bunny entirely.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Simple minimalist layout
- + Appropriate color palette
- − Text is nonsensical gibberish
- − The cloche is missing the bottom plate and steam details
- − Logo elements feel disconnected and unrefined
Stable Diffusion 3.5 Large
- + Excellent text rendering of 'Caffé Florian' and 'Est. 1720'
- + Successfully includes all requested elements: steam, cloche, and banner
- + Professional vintage aesthetic with good use of texture
- − Added an extra 'e' to 'Caffé'
- − The steam inside the cloche looks a bit like flames
Verdict: Stable Diffusion 3.5 Large is the clear winner as it successfully rendered most of the requested text and adhered to every element of the prompt, including the banner and steam. DALL-E 2 failed significantly on text legibility and provided a much more basic, incomplete interpretation of the cloche icon.
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.”
DALL-E 2
Stable Diffusion 3.5 Large
AI Judge Analysis
DALL-E 2
- + Follows the requested color palette well with navy and muted red.
- + Good alignment of blocky infographic text areas.
- − Nonsensical typography even for a base generation.
- − Completely fails to follow the 6-step logical sequence requested.
- − Visuals are chaotic and do not resemble an Apollo 11 mission map.
Stable Diffusion 3.5 Large
- + Much closer to a modern vector infographic style.
- + Successfully categorizes visual elements across the layout.
- + Includes a detailed lunar surface and rocket iconography.
- − Uses a Space Shuttle/Buran style vehicle instead of the requested Saturn V.
- − Text is mostly gibberish despite attempting labels.
- − The layout doesn't clearly display the 6 specific chronological steps requested.
Verdict: Stable Diffusion 3.5 Large is the clear winner here as it produces a coherent, visually appealing poster that actually resembles a space infographic, whereas DALL-E 2 produces a messy, nonsensical collage. While Stable Diffusion 3.5 Large failed to accurately depict the Saturn V (showing a shuttle instead), it followed the stylistic instructions and color palette much more effectively.
Next steps
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