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
FLUX.2 [dev]
24.3
arena score
#19 of 62 in Text-to-Image
Vote tally
Where the votes landed
DALL-E 2
0.0%
win rate
Ties
0.0%
FLUX.2 [dev]
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
FLUX.2 [dev]
0% wins
0% ties
100% wins
AI Judge Analysis
DALL-E 2
- + Decent glass reflection on the table surface.
- − Failed almost every prompt instruction.
- − The blue sphere is missing, replaced by a giant blue pot.
- − The red book is missing, replaced by a red core inside the cube.
- − Poor image clarity and scale.
FLUX.2 [dev]
- + Perfect adherence to all spatial and object requirements.
- + High photorealism with realistic glass refraction and soft lighting.
- + Excellent composition and depth of field.
- − The window light is slightly more direct than 'soft', though still pleasant.
Verdict: FLUX.2 [dev] followed every detail of the prompt accurately, successfully placing the blue sphere inside the cube and the red book on top. DALL-E 2 failed significantly, confusing the colors and positions of the objects, resulting in a giant blue pot in the background and no book at all.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Successfully captures reflections on wet pavement
- + Achieves an extremely shallow depth of field
- − The subject and bicycle are completely out of focus, losing all detail
- − Fails to show that the man is elderly or Japanese due to blur
- − Low resolution and lacks the requested cinematic realism
FLUX.2 [dev]
- + Excellent adherence to all prompt details, including the elderly Japanese man and the red bicycle
- + Highly realistic skin textures and wet weather effects
- + Perfectly executes the motion blur of passing cars in the background
- − The framing is quite centered, missing the 'imperfect framing' request slightly
Verdict: FLUX.2 [dev] followed every aspect of the prompt with high fidelity, creating a poignant and technically impressive photo with realistic textures and lighting. DALL-E 2 struggled significantly, producing a muddy, out-of-focus image where the primary subjects are unrecognizable, failing the core requirements of the prompt.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Features a very close-up, abstract framing that emphasizes the worn metal texture.
- − Extremely messy execution with significant digital artifacts and poor clarity.
- − Fails most prompt requirements including lifelike eyes, braided hair with beads, and leather textures.
- − Anatomical features are unrecognizable and distorted.
FLUX.2 [dev]
- + Perfect adherence to all prompt elements including braided hair with beads, scars, and ornate engraving.
- + Exceptional photographic clarity with lifelike eyes and realistic skin textures.
- + Excellent lighting effects with the torch reflection and atmospheric bokeh sparks.
- − The leather straps overlap the plate armor in a way that slightly hides the 'close portrait' facial focus.
Verdict: FLUX.2 [dev] significantly outperforms DALL-E 2 by providing a high-fidelity, photorealistic image that follows every detail of the complex prompt. DALL-E 2 produced a low-resolution, distorted image where the subject's face is barely identifiable and most specific details like the beaded braids are completely missing.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Strong bold sans-serif typography
- + High contrast minimal aesthetic
- − Nonsensical text and illegible small print
- − Does not follow requested grid layout or sections
- − Food photos are fragmented and abstract rather than professional
FLUX.2 [dev]
- + Excellent adherence to all layout prompts including sections and grids
- + High visual quality and realism in food photography
- + Clean professional typography with clear pricing and descriptions
- − Minor spelling errors in headers (e.g. 'PIZZAU')
- − Vibrant accents are somewhat repetitive horizontal bars
Verdict: FLUX.2 [dev] significantly outperforms DALL-E 2 by providing a functional, professional menu layout that follows every instruction, including specific sections and a logical grid. While DALL-E 2 produced an abstract artistic piece with gibberish text, FLUX.2 [dev] created a realistic design that looks like a genuine casual dining menu with clear, recognizable food photos.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Strong integration of fiery lighting and atmosphere into the food itself.
- + Captures a sense of violent motion and heat.
- − Text is severely misspelled and barely legible.
- − Poor image clarity with significant artifacts and lack of detail on ingredients.
- − The 'exploded' burger is messy and unappealing.
FLUX.2 [dev]
- + Perfect text rendering for all requested strings, including symbols.
- + Exceptional photorealistic detail on burger ingredients and sauce droplets.
- + Follows all layout requirements including the starburst for the price.
- − The burger components are a bit too clean and static compared to a true 'exploded' view.
- − Composition of ingredients is slightly repetitive near the bottom bun.
Verdict: FLUX.2 [dev] successfully followed every detail of the prompt, delivering high-quality photorealistic textures and perfect text rendering. DALL-E 2 struggled significantly with text legibility and image coherence, resulting in a messy and unappealing advertisement. FLUX.2 [dev] is the clear winner for its professional layout and clarity.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + The texture of the chalk strokes appears authentic for physical chalk.
- − The text is completely illegible gibberish.
- − It fails to follow any of the specific prompt details regarding the menu items or date.
- − The composition is cramped and lacks the requested café background.
FLUX.2 [dev]
- + Excellent prompt adherence with near-perfect text rendering of the requested menu items.
- + Realistic chalkboard aesthetics including smudges and authentic handwriting variations.
- + High-quality composition that includes the cozy café background context.
- − Slight overlapping artifacts in the 'Grilled Octopus' section where some letters appear faintly rewritten.
- − Minor smudge artifacts on the bottom right of the board.
Verdict: FLUX.2 [dev] followed the prompt with remarkable accuracy, rendering almost all the requested text exactly as specified in a beautiful, handwritten chalk style. DALL-E 2 completely failed the text-to-image challenge, producing illegible symbols that do not resemble the requested menu content.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
DALL-E 2
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Strong surreal atmosphere
- + Cleverly reverses the rider dynamic as requested in the prompt
- − Very low image resolution and clarity
- − Anatomical distortions in the horse's legs
- − Muddled textures and lack of cinematic lighting
FLUX.2 [dev]
- + Exceptional visual quality and resolution
- + Cinematic lighting and composition
- + High level of detail on the space suit and horse's mane
- − Completely failed the negative constraint to have the horse on top of the astronaut
Verdict: This comparison highlights the classic trade-off between prompt adherence and image quality. DALL-E 2 successfully followed the difficult 'horse on top' instruction but produced a low-fidelity, distorted image, whereas FLUX.2 [dev] ignored the specific structural constraint but created a stunning, high-definition cinematic scene. Although FLUX.2 failed the logic of the prompt, its overwhelming superiority in quality makes it the more usable image for general purposes.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Attempts the low-light night atmosphere of a city street.
- − Extreme anatomical distortions in the faces and hands of both characters.
- − Fails almost all prompt instructions including the capybara's appearance and the woman's seat position.
- − Low image resolution and significant digital artifacts.
FLUX.2 [dev]
- + Accurately depicts all requested elements including the taxi cap, dark jacket, and paws on the wheel.
- + Excellent photorealistic quality with high-end lighting and texture details on the capybara's fur.
- + Perfectly captures the 'bored' expression of the passenger in the background.
- − The woman appears slightly smaller in scale relative to the interior than she would be in reality.
Verdict: FLUX.2 [dev] followed every instruction perfectly, creating a high-fidelity image that realistically blends the surreal prompt with a believable New York taxi interior. DALL-E 2 failed significantly, producing a distorted image with unrecognizable characters and poor composition.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Successfully captures a weathered, vintage parchment aesthetic.
- + The ornate border has a complex, hand-drawn feel.
- − Text is largely illegible with incorrect spellings like 'PAARTY' and 'NIGHTT'.
- − Fails to include specific required details like the date, time, and location.
- − Lacks a central jack-o'-lantern as requested.
FLUX.2 [dev]
- + Excellent typography with perfect adherence to the requested text and event details.
- + Clean, polished composition with high-quality rendering of webs, thorns, and the jack-o'-lantern.
- + Accurately represents all prompt elements including the banner and specific background details.
- − The aesthetics are slightly more modern-digital than 'vintage gothic'.
- − The scroll banner is small and placed very low in the frame.
Verdict: FLUX.2 [dev] is the clear winner as it followed every instruction, including the specific date and location text which DALL-E 2 completely failed to render. DALL-E 2 produced an atmospheric image, but the nonsensical text and missing central elements make it unusable as an invitation compared to the clean, accurate layout of FLUX.2 [dev].
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Successfully used a light blue background.
- + High resolution image with sharp shadows.
- − Failed significantly on text rendering, showing 'Sush' instead of the requested layout.
- − Failed to include 'JAPAN' text or the flag icon.
- − The 'miniature' 3D aesthetic is weak, appearing more like abstract shapes than sushi.
FLUX.2 [dev]
- + Perfect adherence to all text instructions including 'JAPAN', 'SUSHI', and the flag icon.
- + Excellent miniature 3D aesthetic with high-quality PBR textures for the rice and fish.
- + Correct isometric diorama composition as requested.
- − The lighting is a bit flat compared to the requested 'gentle lighting' which often implies more dynamic highlights.
Verdict: FLUX.2 [dev] followed every single instruction in the prompt, including complex text placement and specific icon requests, whereas DALL-E 2 failed on text, icon, and the general subject matter. FLUX.2 [dev] produced a high-quality 3D diorama that looks exactly like the intended 'miniature cartoon scene'.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Dynamic sense of movement and energy
- + Displays butterflies as requested
- − Severe anatomical distortions in the background animals
- − Blurry textures and low resolution compared to modern standards
- − Poor blending and messy edges on the butterfly and fur
FLUX.2 [dev]
- + Expertly rendered textures and realistic fur details
- + Perfectly captures the golden hour lighting, god rays, and dew sparkles
- + All requested animals are present with distinct, high-quality features
- − The characters are more static/posing than 'tumbling' as requested
Verdict: FLUX.2 [dev] significantly outperforms DALL-E 2 in every technical metric, providing a high-fidelity image with realistic fur, lighting, and anatomy. DALL-E 2 struggled with the complexity of multiple animals, resulting in distorted, unrecognizable shapes in the background and a generally low-quality, painterly 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.”
DALL-E 2
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Matches the requested color palette well
- + Captures a minimalist vector-style icon
- − Text is nonsensical and does not follow the prompt
- − Missing the steam and banner elements
- − Very low technical detail and presence of artifacts
FLUX.2 [dev]
- + Perfect text rendering for both the name and the established date
- + Completely adheres to all requested elements including cloche, steam, and banner
- + Professional vector aesthetic with subtle background texture
- − Minimalist style is slightly more 'detailed' than 'minimalist' but still fits the vintage vibe
Verdict: FLUX.2 [dev] successfully followed every instruction in the prompt, producing a clean, professional logo with perfect typography and all requested visual elements. DALL-E 2 failed significantly on text rendering and omitted several key components like the banner and the specific name requested.
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
FLUX.2 [dev]
AI Judge Analysis
DALL-E 2
- + Adheres strictly to the requested NASA-inspired color palette.
- + Achieves a complex technical layout that feels like an authentic schematic.
- − Text is completely illegible and contains severe misspellings like 'ALLPOO'.
- − Fails to clearly represent the specific 6-step sequence requested in the prompt.
- − Contains messy, incoherent graphical artifacts.
FLUX.2 [dev]
- + Excellent adherence to the requested 6-step sequence with clear icons.
- + Clean, modern vector aesthetic with high-quality illustrations.
- + Correctly renders 'Apollo 11 Mission' and the names of the astronauts.
- − Includes some redundant or garbled labels beneath certain icons.
- − The layout is a bit repetitive, resembling a grid of stickers more than a cohesive infographic poster.
Verdict: FLUX.2 [dev] significantly outperforms DALL-E 2 by accurately following the multi-step instructional prompt and generating legible, relevant iconography. While DALL-E 2 captures a more complex aesthetic, its failure to generate readable text ('ALLPOO') and its inability to represent the specific mission steps make it unusable for an infographic task.
Next steps
Explore each model
Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment