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 Medium
16.8
arena score
#58 of 62 in Text-to-Image
Vote tally
Where the votes landed
DALL-E 2
50.0%
win rate
Ties
0.0%
Stable Diffusion 3.5 Medium
50.0%
win rate
50.0%
0.0% ties
50.0%
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 Medium
AI Judge Analysis
DALL-E 2
- + Features a wooden table and a glass cube.
- + Captures soft lighting effectively.
- − Failed to place a red book on top (red color was placed inside).
- − Failed to place a blue sphere inside (the background pot is blue instead).
- − Poor compositional scale between the cube and background elements.
Stable Diffusion 3.5 Medium
- + Excellent prompt adherence for all spatial relationships.
- + Realistic glass textures and reflections.
- + Accurate interpretation of the soft window light from the left.
- − The blue sphere appears to be floating without a physical support inside the cube.
Verdict: Stable Diffusion 3.5 Medium followed every instruction in the prompt, correctly placing the red book on top, the blue sphere inside, and the plant behind the cube. DALL-E 2 struggled significantly with spatial logic, mixing up the colors of the objects and failing to include the book entirely.
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 Medium
AI Judge Analysis
DALL-E 2
- + Successfully captures reflections on wet pavement
- + Delivers an 'imperfect framing' look as requested
- − The main subject is completely out of focus, failing the '50mm' and 'natural skin texture' requirements
- − Low resolution and lacks any discernable detail of the elderly man
Stable Diffusion 3.5 Medium
- + Excellent adherence to the elderly man subject and red bicycle
- + Highly realistic skin textures and lighting
- + Captures the rain and reflections with high clarity
- − Lacks significant motion blur from passing cars
- − The bicycle geometry is slightly warped near the pedals
Verdict: Stable Diffusion 3.5 Medium far outperforms DALL-E 2 by providing a clear, high-quality image that respects the detailed descriptions of the elderly man and his environment. DALL-E 2 misinterpreted the request for shallow depth of field and imperfect framing by blurring the primary subject, rendering the prompt's specific textural details impossible to see.
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 Medium
AI Judge Analysis
DALL-E 2
- + Features a very close-up, gritty texture
- − Low resolution and blurry quality
- − Fails to show life-like eyes or braided hair or many requested details clearly
- − Anatomically confusing and lacks the requested ornate aesthetics
Stable Diffusion 3.5 Medium
- + Excellent adherence to all prompt details including braided hair, beaded accents, and ornate armor engravings
- + High visual clarity with realistic skin textures and lifelike eyes
- + Beautiful lighting effects with warm torchlight and bokeh sparks
- − The braids appear slightly digital and overly uniform at the scalp
Verdict: Stable Diffusion 3.5 Medium successfully interprets every aspect of the prompt, creating a high-fidelity, cinematic portrait with rich detail in the armor and facial features. DALL-E 2 produces a low-quality, abstracted image that fails to represent most of the descriptive elements requested.
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 Medium
AI Judge Analysis
DALL-E 2
- + Strong bold typography that feels modern and impactful.
- + The food images are abstracted in an artistically interesting geometric layout.
- − Fails to follow the 'grid' structure requested for a menu.
- − Text is completely illegible gibberish and cuts off the right side of the page.
- − Lacks the required sections for appetizers, pizza, and mains.
Stable Diffusion 3.5 Medium
- + Perfectly adheres to the grid layout with multiple food photos.
- + Successfully includes categorized sections and price lists as requested.
- + Clean, professional composition that accurately reflects a casual dining menu.
- − Text is slightly distorted and contains misspellings.
- − The fonts used are more whimsical than 'bold sans-serif'.
Verdict: Stable Diffusion 3.5 Medium is the clear winner as it accurately interprets the prompt for a functional menu layout including grids, price sections, and food categorization. DALL-E 2 produces a more abstract artistic spread that fails to function as a readable menu or follow the specific layout instructions.
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 Medium
AI Judge Analysis
DALL-E 2
- + Captures the 'exploded' and 'dynamic' motion requested in the prompt.
- + Atmospheric fiery effects and glowing embers are well-executed.
- − Poor text rendering with spelling errors like 'MARGIC' and 'BAGUEC'.
- − Lack of photorealistic detail, appearing more like an abstract digital painting.
- − Failed to include the secondary message and starburst price.
Stable Diffusion 3.5 Medium
- + Excellent text rendering with high accuracy for all requested phrases.
- + Very high photorealistic detail and sharp resolution.
- + Features a clean starburst element with the correct price.
- − Failed to provide the 'exploded' view where components are suspended separately in mid-air.
- − The fiery background is a bit generic compared to the requested glowing ember effect.
Verdict: Stable Diffusion 3.5 Medium produces a much more professional and legible advertisement with perfect text rendering, though it fails to explode the burger components as requested. DALL-E 2 attempts the complex composition described but fails significantly on image quality and text accuracy. Stable Diffusion 3.5 Medium is the preferred choice for a usable marketing asset.
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 Medium
AI Judge Analysis
DALL-E 2
- + The text has a realistic, messy chalk texture.
- − Completely failed to render any legible text from the prompt.
- − The text is just nonsensical scribbles.
- − The image is low resolution and poorly framed.
Stable Diffusion 3.5 Medium
- + Successfully rendered several keywords like 'TODAY SPECIALS', 'Mushroom', and 'Cookies'.
- + The composition is excellent, showing a framed chalkboard in a café setting.
- + High visual quality with realistic wood textures and chalk smudges.
- − Failed to render the full phrases and dates accurately.
- − Spelling errors are present in almost every word (e.g., 'Musrroom', 'Oocclutels').
- − The handwriting style changes between printed and cursive, contrary to the 'exact same style' instruction.
Verdict: Stable Diffusion 3.5 Medium is the clear winner as it demonstrates a much higher level of prompt understanding and visual fidelity compared to DALL-E 2. While Stable Diffusion 3.5 Medium struggles with complex spelling and specific date formatting, it produces a coherent and aesthetically pleasing image, whereas DALL-E 2 produces illegible gibberish.
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 Medium
100% wins
0% ties
0% wins
AI Judge Analysis
DALL-E 2
- + Attempts a painterly, surreal texture
- + Features a dramatic composition with a planet in the background
- − Failed the negative constraint: the astronaut is riding the horse
- − Significant anatomical distortions on the horse and astronaut
- − Low resolution and grainy textures
Stable Diffusion 3.5 Medium
- + High resolution with crisp details on the spacesuit and horse
- + Clean, cinematic lighting and vibrant colors
- + Captures a sense of scale with a detailed planet below
- − Failed the negative constraint: the astronaut is riding the horse
- − Anatomical issues with the horse's legs appearing elongated and segmented
Verdict: Both DALL-E 2 and Stable Diffusion 3.5 Medium failed the spatial reasoning challenge to place 'horse on top' of the astronaut, providing a standard astronaut-riding-horse image instead. Stable Diffusion 3.5 Medium is the winner due to significantly better image quality, lighting, and detail despite the prompt adherence failure shared by both models.
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 Medium
AI Judge Analysis
DALL-E 2
- + Successfully includes the phone mentioned in the prompt.
- − Extreme anatomical distortions in the human face and hands
- − The capybara is unrecognizable and looks like a distorted yellow mass
- − Composition is chaotic and does not follow the interior/exterior logic of a car
Stable Diffusion 3.5 Medium
- + High visual quality with realistic fur and lighting
- + Strong adherence to the prompt including the capybara's outfit and expression
- + Effective cinematic composition with blurred Manhattan lights in the background
- − The passenger is not holding a phone as requested
- − Minor blurring issues on the human's facial features in the background
Verdict: Stable Diffusion 3.5 Medium produced a high-quality, professional-looking image that followed nearly all prompt instructions, including the specific hat and jacket for the capybara. DALL-E 2 struggled significantly with this complex scene, resulting in nightmare-like anatomical glitches and a failure to render a recognizable capybara.
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 Medium
0% wins
0% ties
100% wins
AI Judge Analysis
DALL-E 2
- + Captures a strong vintage gothic aesthetic.
- + Expressive, painterly art style.
- + The border design integrates well with the theme.
- − Text is completely illegible and nonsensical.
- − Failed to include specific event details requested in the prompt.
- − Lacks a clear central jack-o-lantern.
Stable Diffusion 3.5 Medium
- + Excellent adherence to all visual elements like jack-o-lanterns, webs, and trees.
- + High image clarity and professional layout.
- + Text is mostly legible even with minor misspellings.
- − Multiple spelling errors in the primary text (e.g., 'Halloweeen', 'Inviloween', 'Loccation').
- − Text layout is a bit plain compared to the 'elegant gothic' request.
Verdict: Stable Diffusion 3.5 Medium is the clear winner as it successfully incorporated almost every specific element requested, including the jack-o-lanterns and the event details. While it struggled with perfect spelling, DALL-E 2 produced completely garbled text and failed to follow the specific content requirements of the prompt.
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 Medium
AI Judge Analysis
DALL-E 2
- + Clean isometric perspective
- + Distinct 3D shadows and lighting
- − Incomplete text rendering showing 'Sush' instead of 'SUSHI'
- − Failed to include 'JAPAN' text or flag icon
- − The sushi models look like abstract plastic shapes rather than appetizing food
Stable Diffusion 3.5 Medium
- + Accurately rendered text for both 'JAPAN' and 'SUSHI'
- + Beautiful material rendering on the ikura (salmon roe) and seaweed
- + Followed the composition instructions perfectly, including the diorama base
- − Missed the flag icon requirement
- − The 'S' in SUSHI has a slight transparency artifact
Verdict: Stable Diffusion 3.5 Medium significantly outperformed DALL-E 2 by adhering to almost all prompt instructions, including complex text rendering and material quality. While DALL-E 2 produced a clean clinical look, it failed on key text elements and the 3D models lacked the 'soft refined textures' and realism present in the Stable Diffusion output.
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 Medium
AI Judge Analysis
DALL-E 2
- + Attempts to show the animals in active, playful motion as requested.
- − Severely lacks the requested 'hyper-photorealistic' quality, appearing like a low-resolution collage.
- − Contains significant anatomical artifacts, such as the kitten's distorted limbs and the unrecognizable fox/bunny shape.
- − The butterflies are poorly rendered and look like flat clip-art.
Stable Diffusion 3.5 Medium
- + Excellent adherence to the 'golden sunrise' and 'god rays' lighting request.
- + High visual quality with detailed fur textures and expressive eyes.
- + Cohesive and vibrant composition that fits the 'wholesome' aesthetic.
- − Missing the baby bunny requested in the prompt.
- − The 'tabby kitten' looks somewhat fox-like or hybrid in appearance.
- − The animals are more static/posing than 'tumbling' and 'chasing'.
Verdict: Stable Diffusion 3.5 Medium is the clear winner due to its superior image quality and lighting, despite missing one of the four animals. DALL-E 2 produced a very low-quality image with significant anatomical distortions and a dated, collage-like feel that failed the 'hyper-photorealistic' requirement.
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 Medium
AI Judge Analysis
DALL-E 2
- + Simple minimalist layout
- + Follows the color palette accurately
- − Nonsensical text and gibberish lettering
- − Poorly defined and messy steam graphics
- − Imbalanced composition
Stable Diffusion 3.5 Medium
- + Excellent vintage illustration style
- + Text is mostly legible and correctly formatted
- + Complex and balanced vector emblem composition
- − Minor spelling errors (Caffee Florrian, Est 170)
- − Less minimalist than requested
- − Banner text at the bottom is illegible
Verdict: Stable Diffusion 3.5 Medium produced a much higher quality logo with a sophisticated vintage aesthetic and recognizable text elements, despite small typos. DALL-E 2 failed significantly on prompt adherence regarding the specific text and produced a messy, incoherent design with distorted characters.
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 Medium
AI Judge Analysis
DALL-E 2
- + Matches the requested NASA-inspired color palette perfectly.
- + Captures the technical, complex look of a data infographic.
- − Text is nonsensical and includes unfortunate misspellings of the core subject.
- − Fails to follow the chronological 6-step logical flow requested in the prompt.
- − The layout is cluttered and chaotic rather than 'clean and modern'.
Stable Diffusion 3.5 Medium
- + Successfully follows the sequential step-by-step infographic layout requested.
- + Maintains better legibility of the title and various labels.
- + The vector style is cleaner and align better with the requested flat design.
- − The icons do not strictly match the specific descriptions (e.g., Earth and Moon icons are confused/repetitive).
- − Text still contains significant spelling errors and gibberish.
- − The composition feels slightly empty on the right side.
Verdict: Stable Diffusion 3.5 Medium is the winner because it actually attempts to create the requested 6-step sequential infographic structure, whereas DALL-E 2 produces a chaotic jumble of technical-looking elements with no logical flow. While neither model achieved perfect text or iconography, Stable Diffusion 3.5 Medium's adherence to the prompt's layout and themes makes it far more useful as a starting point.
Next steps
Explore each model
Stability AI's 2.5-billion parameter Multimodal Diffusion Transformer with improvements (MMDiT-X) text-to-image model optimized for consumer hardware, featuring improved image quality, typography, and complex prompt understanding