Head to head
Esc

Models · slot A

to navigate to pick

FLUX1.1 [pro] Ultra Black Forest Labs Stable Diffusion 3.5 Large Stability AI

Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.

FLUX1.1 [pro] Ultra

17.9 arena score

#51 of 62 in Text-to-Image

Skill signature · Text-to-Image

Stable Diffusion 3.5 Large

22.9 arena score

#30 of 62 in Text-to-Image

Vote tally

Where the votes landed

FLUX1.1 [pro] Ultra

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Perfect adherence to all spatial instructions in the prompt.
  • + Highly realistic textures on the book cover and the glass edges.
  • + Excellent rendering of the plant's appearance through the glass diffraction.
  • The sphere is floating unnaturally without a visible base, which might be a stylistic choice but lacks physical realism.

Stable Diffusion 3.5 Large

  • + Realistic lighting and caustic effects from the sphere onto the book.
  • + Clear, sharp glass edges.
  • Failed the primary spatial prompt by placing the book inside the cube instead of on top.
  • The sphere is resting on the book rather than just being 'inside the cube' independently.
  • The background is cluttered with many extra books and furniture not requested.

Verdict: FLUX1.1 [pro] Ultra followed every instruction perfectly, placing the red book on top of the cube and the plant behind it. Stable Diffusion 3.5 Large failed on the spatial relationships, placing the book inside the cube and adding various unrequested background elements.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent execution of long-exposure motion blur on the passing vehicle.
  • + Superior rendering of complex reflections on wet asphalt.
  • + High photographic clarity and realistic skin tones.
  • The man appears to be leaning on the bike rather than actively repairing it.
  • The framing feels a bit too centered for an 'imperfect' candid request.

Stable Diffusion 3.5 Large

  • + Successfully captures the 'repairing' action with a更加 involved pose.
  • + Captures the atmosphere of 'light rain' more visibly with falling droplets.
  • + Follows the 'imperfect framing' prompt better with an off-center subject.
  • Missed the 'motion blur' requirement for the passing car.
  • Lower overall resolution and slightly mushy textures in the background.
  • Anatomical issues with the man's feet and the bicycle pedal structure.

Verdict: FLUX1.1 [pro] Ultra produces a much more technically impressive image with realistic motion blur and sophisticated reflections, though it lacks the specific action of 'repairing'. Stable Diffusion 3.5 Large captures the scene's mood and the physical activity of the man better, but fails the motion blur requirement and has more AI artifacts. FLUX1.1 [pro] Ultra is the winner for its convincing photographic quality and adherence to complex technical camera prompts.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Perfect visual representation of 'battle-worn' with intense grime and realistic skin texture
  • + Exceptional cinematic lighting with strong torchlight highlights and lens flares
  • + Superior macro-detail on the leather and fabric textures
  • The 'beads' in the hair are very subtle and almost meld with the hair strands
  • Extremely shallow depth of field obscures the back half of the head

Stable Diffusion 3.5 Large

  • + Excellent engraving detail on the pauldrons and chestplate
  • + Very clear depiction of the braided hairstyle
  • + Good balance of environment/background bokeh while keeping the subject sharp
  • The skin and hair look a bit too clean/conditioned for a 'battle-worn' character
  • Missing the requested 'beads' in the braids
  • Lighting feels more like daylight with secondary orange hits rather than a primary warm torchlight source

Verdict: FLUX1.1 [pro] Ultra better captures the gritty, atmospheric essence of the prompt with realistic skin imperfections and superior cinematic lighting. While Stable Diffusion 3.5 Large offers impressive armor engravings, it fails to include the requested beads and feels less 'battle-worn' than the FLUX output.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent professional mockup layout on a clean white background
  • + Clear headers for sections like Appetizers and Mains
  • + Vibrant and well-lit food photography that looks appetizing
  • Text contains several spelling errors (e.g., 'Pizzans', 'Meere Prer')
  • Small body text is mostly gibberish or garbled characters

Stable Diffusion 3.5 Large

  • + Strong grid-based composition for food photos
  • + Bold, readable sans-serif header fonts
  • + Clean white-space usage in the central column
  • Text rendering is very poor with significant spelling errors in all headers (e.g., 'MAIMAES', 'PNC MALS')
  • The layout feels more like a social media post or poster than a functional restaurant menu

Verdict: FLUX1.1 [pro] Ultra produces a much more realistic and professional menu layout that would be suitable for a design mockup, despite some minor spelling errors. Stable Diffusion 3.5 Large has very poor text legibility and the central menu column contains mostly nonsense characters, yielding a less practical design for a restaurant.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent typography with perfect spelling and the requested glowing effect.
  • + Highly detailed food textures with a clean, professional advertising layout.
  • + Successfully incorporates all requested design elements including the price and secondary message.
  • The burger structure is more of a stack than a chaotic 'exploded' view.
  • Includes some small AI hallucinations like incoherent fine print at the bottom corners.

Stable Diffusion 3.5 Large

  • + Strong sense of 'fire' and heat with a more integrated background.
  • + The floating effect feels dynamic with the sauce droplets.
  • Failed to include any of the requested text elements.
  • Missing the 'exploded' component layering requested in the prompt.
  • Lacks the advertisement composition, looking more like a standard food image.

Verdict: FLUX1.1 [pro] Ultra is the clear winner as it followed every instruction, including the complex text integration and specific price formatting. While Stable Diffusion 3.5 Large produced a visually appealing and fiery image, it completely ignored the text requirements and the 'exploded' burger layout requested in the prompt.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + The handwriting style is very realistic with convincing chalk texture and natural variations.
  • + The year 2026 is rendered correctly as requested.
  • + The layout closely follows a single-column menu format common in cafes.
  • Contains significant spelling errors and repetitive gibberish text in the lower half.
  • Failed to render the title in an 'elegant cursive' style, opting for print instead.

Stable Diffusion 3.5 Large

  • + The broader cafe environment is aesthetically pleasing and well-composed.
  • + The text layout is clean and organized into boxed sections.
  • Failed the date request, rendering 2024 instead of 2026.
  • The text is largely illegible/nonsense ('TODAAY', 'Ottpups', 'Cockies').
  • The handwriting looks more like a digital brush than natural chalk variation.

Verdict: FLUX1.1 [pro] Ultra produced a much more realistic chalk texture and captured the specific year requested, although it struggled with spelling consistency towards the bottom of the board. Stable Diffusion 3.5 Large failed on the specific date and produced mostly gibberish text that does not resemble the requested menu items. FLUX1.1 [pro] Ultra is the clear winner for its superior text rendering and adherence to the specific prompt details.

The Reversed Rodeo

Text-to-Image

“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent clarity and resolution of textures
  • + Dynamic posing with the horse rearing
  • + Clean and crisp background with good planetary depth
  • Failed the negative constraint to have the horse on top
  • The astronaut's interaction with the reins is anatomically inconsistent with his hand placement

Stable Diffusion 3.5 Large

  • + Stronger cinematic atmosphere with star dust and lighting effects
  • + Realistic fabric texture on the astronaut suit
  • + Better sense of scale and motion
  • Failed the negative constraint to have the horse on top
  • The horse's face has slight anatomical distortions around the muzzle

Verdict: Both models failed the specific spatial logic request of having the 'horse on top' of the astronaut, instead defaulting to the standard horse-rider trope. FLUX1.1 [pro] Ultra produces a much sharper and cleaner image, while Stable Diffusion 3.5 Large offers a more atmospheric and painterly aesthetic.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent fur texture and photographic lighting.
  • + Successfully included both the driver and the passenger as requested.
  • + Realistic bokeh effect on the New York street background.
  • The passenger is holding the steering wheel instead of sitting in the back seat.
  • The capybara's head is fused with the car frame and its body proportions are awkward.

Stable Diffusion 3.5 Large

  • + Strong prompt adherence for the capybara's outfit and positioning at the wheel.
  • + Good interior taxi details like the dashboard and door handle.
  • + Creatively interpreted the capybara with human-like clothing layers.
  • Completely failed to include the businesswoman in the back seat.
  • The capybara's anatomy becomes confused near the lap area.
  • The hat logo is nonsensical gibberish.

Verdict: FLUX1.1 [pro] Ultra produced a more photorealistic image with a better cinematic atmosphere, but it struggled with spatial reasoning, placing the passenger in the front seat holding a second steering wheel. Stable Diffusion 3.5 Large followed the capybara's clothing description more accurately but completely ignored the instruction to include a passenger in the back seat. FLUX1.1 [pro] Ultra is the winner as it attempted all elements of the complex prompt and achieved much higher visual fidelity, despite the compositional error.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Most accurate text rendering of all requested details including date, time, and location.
  • + Excellent composition with a focused, glowing jack-o-lantern and clean gothic typography.
  • + Very high visual quality with intricate thorns and a web border that fits the vintage parchment theme.
  • Redundant text repetition in the middle of the scroll (e.g., 'noickt or a night of frights').
  • Missing the word 'Party' in the main heading.

Stable Diffusion 3.5 Large

  • + Dramatic and atmospheric lighting with a large moon and twisted trees.
  • + Great aesthetic interpretation of a 'dark parchment' with torn, weathered edges.
  • + Includes all words in the main title: 'Halloween Party Invitation'.
  • Failed to include the specific event details (Date, Time, Location) at the bottom.
  • Text inside the scroll banner becomes garbled and messy at the end.
  • Composition is a bit cluttered, making the text harder to read against the busy background.

Verdict: FLUX1.1 [pro] Ultra is the clear winner for its superior text rendering and adherence to the layout instructions, successfully including the date, time, and location exactly as requested. While Stable Diffusion 3.5 Large has a more creative and atmospheric artistic style, it completely failed to include the essential event details, making it useless as a functional invitation.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent typography integrated into the background as requested.
  • + Clean, professional miniature aesthetic with soft, refined 3D textures.
  • + Perfect adherence to the isometric perspective and diorama layout.
  • Includes a strange, small insect or bug detail on the wooden board.
  • Rice texture is slightly lumpy rather than individual grains.

Stable Diffusion 3.5 Large

  • + Realistic food textures, especially the salmon and rice grains.
  • + Large variety of sushi types on the platter.
  • Failed to place text as part of the background, instead putting it on a small sign.
  • Perspective is more of a standard 3/4 view rather than a strict 45-degree isometric top-down.
  • The flag is placed in the food rather than as an icon with the text.

Verdict: FLUX1.1 [pro] Ultra followed the complex layout instructions much more accurately, creating a clean graphic design with the text correctly positioned in the background. While Stable Diffusion 3.5 Large had impressive texture work on the food itself, it failed to execute the isometric diorama style and typography requirements properly.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Excellent fur texture and lighting integration on all baby animals.
  • + Very cute, cohesive art style that fits the 'wholesome vibe' perfectly.
  • + Captures all four requested animals clearly with distinct features.
  • The animals are mostly sitting rather than 'playfully chasing' or 'tumbling' as requested.
  • The lighting and proportions feel more like a Pixar CGI render than 'hyper-photorealistic'.

Stable Diffusion 3.5 Large

  • + Better adherence to the 'chasing' and 'tumbling' action in the prompt.
  • + Higher level of realism in the fur and environment textures.
  • + Includes visible 'dew sparkles' and more dynamic movement.
  • The fox kit's face in the background is slightly distorted and less detailed.
  • One butterfly appears to be awkwardly growing out of the rabbit's ear.

Verdict: Stable Diffusion 3.5 Large followed the prompt's action requirements more closely, showing the animals running and chasing rather than posing. While FLUX1.1 [pro] Ultra created a more polished, aesthetically pleasing image with perfect facial features, it failed to capture the dynamic 'tumbling' motion requested, opting for a static portrait.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + High-quality vector illustration style
  • + Excellent use of warm brown and cream tones
  • + Clean, balanced composition suitable for a professional logo
  • Spelling error in the main brand name: 'Framilian' instead of 'Florian'
  • The 'Est. 1720' is not on the banner as requested, but placed below it

Stable Diffusion 3.5 Large

  • + Accurately spelled the brand name 'Florian' (though added an extra 'e' to Caffé)
  • + Includes subtle textures on the background as requested
  • + Follows the banner layout instruction for the text
  • Visual quality is lower with jagged edges on the dome and steam
  • The cloche dome is floating awkwardly and merged with abstract lines
  • Typography is less sophisticated and lacks a premium 'vintage' feel

Verdict: FLUX1.1 [pro] Ultra produces a much more professional and aesthetically pleasing vector emblem, though it fails on the specific spelling of the name. Stable Diffusion 3.5 Large follows the text instruction better but suffers from poor execution of the central graphic and lower overall image clarity. FLUX1.1 [pro] Ultra is the likely winner for its superior design quality, despite the typo.

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.”

FLUX1.1 [pro] Ultra
Stable Diffusion 3.5 Large

AI Judge Analysis

FLUX1.1 [pro] Ultra

  • + Stronger adherence to the flat-vector style requested.
  • + Legible main headings that closely follow the prompt's steps.
  • + Excellent use of the NASA-inspired color palette.
  • Nonsense filler text in the body paragraphs.
  • Logically confusing layout with icons appearing in random locations on the page.

Stable Diffusion 3.5 Large

  • + High level of intricate detail within the technical diagrams.
  • + Consistent vintage-technical aesthetic throughout the composition.
  • Failed to follow the flat-vector style, opting for a complex illustrative look.
  • Included incorrect hardware like a Space Shuttle instead of a Saturn V.
  • Failed to provide the specific 6-step numbered list requested.

Verdict: FLUX1.1 [pro] Ultra is the clear winner for its adherence to the requested flat-vector style and its ability to render the specific steps and headings requested in the prompt. While Stable Diffusion 3.5 Large created a visually complex image, it ignored the stylistic constraints and failed to accurately depict the historical Apollo hardware, showing a shuttle-like vehicle instead.

Next steps

Explore each model