Black Forest Labs' 12-billion parameter multimodal flow transformer for in-context image generation and editing with character consistency, typography handling, and commercial-ready quality
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 Kontext [pro]
20.3
arena score
#41 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
FLUX.1 Kontext [pro]
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent adherence to spatial instructions with the book placed on top as requested.
- + Highly realistic photographic quality with natural lighting and depth of field.
- + Accurate representation of glass refractions and plant visibility through the cube.
Stable Diffusion 3.5 Large Turbo
- + Sharp, clean rendering of the wooden table and sphere texture.
- + Vibrant colors and high contrast.
- − Failed the primary spatial prompt by placing the book inside the cube instead of on top.
- − The plant is positioned to the side rather than strictly behind the cube.
- − The sphere is partially clipping into the book.
Verdict: FLUX.1 Kontext [pro] followed every part of the prompt perfectly, including the specific spatial arrangement of the red book on top of the cube. Stable Diffusion 3.5 Large Turbo failed to place the book correctly and suffered from minor clipping issues between the objects inside the cube.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent skin texture and realistic lighting
- + Captures the mood of light rain and street reflections perfectly
- + Achieves an authentic 'candid' feel with a shallow depth of field
- − The subject appears to be more riding or holding the bike than actively 'repairing' it
- − Missing the specific prompt request for motion blur from passing cars
Stable Diffusion 3.5 Large Turbo
- + Clearly shows the act of repairing or inspecting the bike
- + Follows the request for a red bicycle and wet pavement
- − Anatomical failure where the head/hat blends into the shoulder
- − Lacks the requested 50mm shallow depth of field, looking very flat
- − The rain effect is rendered as unrealistic vertical white lines
Verdict: FLUX.1 Kontext [pro] creates a much more believable and high-quality image with realistic skin textures and atmospheric lighting that perfectly captures the requested 'candid street photo' aesthetic. While Stable Diffusion 3.5 Large Turbo attempts the specific 'repairing' action more directly, it suffers from significant anatomical distortions and a lack of depth that makes it look like a lower-quality digital generation.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent photorealism with skin texture and natural facial hair.
- + Subtle, realistic engraving on the plate armor that looks historical.
- + Soft, cinematic lighting that feels integrated into the scene.
- − The beads in the hair are very subtle and barely visible.
- − The 'bokeh sparks' are a bit sparse compared to the prompt's request.
Stable Diffusion 3.5 Large Turbo
- + Ornate armor design with high contrast and decorative detail.
- + Bold braids that clearly feature a bead-like fastener.
- + Striking use of color and lighting.
- − The 'scars' look more like smeared paint or blood than healed skin.
- − Hyper-saturated, plastic-like texture on the skin and hair lacks realism.
- − The earring and hair structure contain slight anatomical inconsistencies.
Verdict: FLUX.1 Kontext [pro] provides a significantly more realistic and high-quality image, capturing the 'battle-worn' aesthetic through subtle skin textures and believable lighting. Stable Diffusion 3.5 Large Turbo produces a more stylized, almost comic-book-like appearance with skin that looks airbrushed and 'dirt' that looks like face paint.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Strictly follows the layout request with clear sections for Appetizers, Pizza, and Mains.
- + Excellent typography with bold sans-serif fonts and legible (though nonsensical) text.
- + Cohesive minimalist aesthetic that genuinely looks like a professional menu design.
- − The food items listed under the sections do not always match the headers (e.g., 'Pizza' items listed under 'Mains').
- − Text is gibberish despite having a realistic appearance.
Stable Diffusion 3.5 Large Turbo
- + Includes a grid of colorful food photos as requested in the prompt.
- + High vibrancy and saturation in the food imagery.
- − Fails to integrate the text and images into a single professional menu layout, instead creating a scattered collage.
- − The text is unreadable and poorly rendered with 'Mians' as a typo for 'Mains'.
- − Composition is cluttered and does not feel like a functional graphic design piece.
Verdict: FLUX.1 Kontext [pro] produced a highly professional and usable menu layout that perfectly captured the 'modern minimalist' aesthetic requested. In contrast, Stable Diffusion 3.5 Large Turbo created a chaotic collection of food assets and poorly rendered text blocks that failed to form a coherent graphic design.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent typography rendering with perfect spelling and fiery effects.
- + High photorealistic detail in the food textures and fire embers.
- + Better adherence to the 'exploded' aspect of the prompt with separation between components.
- − The price element is repeated twice, which was not requested.
- − Missing the starburst element for the price.
Stable Diffusion 3.5 Large Turbo
- + Dynamic fire and smoke effects.
- + Vibrant colors and high contrast.
- − Failed to include any of the requested text.
- − Missing the 'exploded' suspended components; the burger appears mostly stacked.
- − Less photorealistic and more illustrative/CG in style.
Verdict: FLUX.1 Kontext [pro] is the clear winner as it followed all complex text instructions perfectly and achieved a much higher level of photorealism. Stable Diffusion 3.5 Large Turbo failed to include any of the required text and did not effectively represent the 'exploded' burger concept.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent text rendering with nearly perfect spelling and legible grammar.
- + Followed the prompt instructions for specific dates and menu items accurately.
- + Highly realistic chalk texture with grainy strokes and natural handwriting variations.
- − The title is in a handwriting style but not specifically 'elegant cursive' as requested.
- − A small typo 'foiur' in the additional footer text.
Stable Diffusion 3.5 Large Turbo
- + Successfully used a cursive/script font for the header as requested.
- + Pleasant composition with a wide-shot view of the café environment.
- − Poor text rendering with numerous spelling errors and garbled words like 'trulale' and 'ocotpg'.
- − Failed to include the specific year (2026) and specific prices ($24) correctly.
- − Text looks like a digital brush rather than a realistic chalk texture.
Verdict: FLUX.1 Kontext [pro] is the clear winner as it successfully rendered the complex menu items and dates with high legibility and a convincing chalk texture. Stable Diffusion 3.5 Large Turbo struggled significantly with the text, producing several nonsensical words and failing to adhere to the specific price points and item descriptions provided in the prompt.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Perfectly follows the specific instruction of 'horse on top' of the astronaut.
- + High level of cinematic detail in the spacesuit textures and lighting.
- + Creative surrealist approach with a smaller horse-riding-astronaut on top of a larger astronaut.
- − Anatomical anomaly where the large astronaut has hooves instead of boots.
Stable Diffusion 3.5 Large Turbo
- + Clear, high-contrast composition with a clean aesthetic.
- + Correct leather-like textures on the saddle and harness.
- − Completely failed the negative constraint/specific directional prompt by putting the astronaut on top.
- − Severe anatomical artifacts on the horse's front legs.
- − Low-quality rendering of the planet and space background.
Verdict: FLUX.1 Kontext [pro] successfully interpreted a complex and unusual prompt instruction ('horse on top, not vice versa'), which requires a level of spatial reasoning that many models lack. Stable Diffusion 3.5 Large Turbo defaulted to a standard trope and ignored the specific prompt details, while also suffering from significant anatomical errors in the horse's legs.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent photorealism with realistic fur texture and lighting.
- + Strong adherence to the requested 'calm, professional expression' of the capybara.
- + The passenger in the background is clearly interacting with a phone as requested.
- − The passenger's hands look slightly distorted near the face.
- − Perspective makes it slightly ambiguous if both paws are on the steering wheel.
Stable Diffusion 3.5 Large Turbo
- + Clearer depiction of both front paws on the steering wheel.
- + The jacket and cap are well-defined and fit the prompt.
- + Vibrant bokeh in the background effectively suggests city lights.
- − The passenger is not looking at her phone, missing a key part of the prompt.
- − The capybara's face has a somewhat artificial, smooth texture compared to Model A.
- − The steering wheel appears to be floating or not connected to a dash.
Verdict: FLUX.1 Kontext [pro] is the winner due to its superior photorealism and better adherence to the specific actions requested in the prompt, particularly the passenger looking at her phone. While Stable Diffusion 3.5 Large Turbo accurately places the capybara's paws on the wheel, it fails on the passenger's details and has a less convincing 'real-world' lighting quality.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent adherence to all text requirements including the specific date and location.
- + High-quality 'dark parchment' aesthetic with moody, cinematic lighting.
- + Detailed and elegant gothic border that fits the theme perfectly.
- − Included a line of hallucinated text ('Your: Vorkleat: Iight & Spans') between the date and time.
Stable Diffusion 3.5 Large Turbo
- + Clean vector-style border with thorns and webs as requested.
- + Good central jack-o-lantern illustration.
- − Failed to include the majority of the requested text, including the scroll banner and event details.
- − The 'moody night sky' and 'dark parchment' textures are very flat and lack atmosphere.
- − The typography is basic and not 'elegant gothic' as specified.
Verdict: FLUX.1 Kontext [pro] is the clear winner as it successfully rendered almost all the complex text requirements, including the specific date, location, and title. While it added one line of gibberish, Stable Diffusion 3.5 Large Turbo failed to include any of the bottom event details or the scroll banner, resulting in a much less useful image for an 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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Perfect text rendering for both 'JAPAN' and 'SUSHI'.
- + Excellent adherence to the 'soft refined textures' and 'gentle lighting' instructions.
- + Clean, minimalist composition that looks like professional 3D graphic design.
- − The rice grains are a bit large and round, resembling pearls more than sushi rice.
- − The flag icon is stylized but slightly undersized compared to the text.
Stable Diffusion 3.5 Large Turbo
- + Accurate isometric perspective and diorama-style layout.
- + The textures on the salmon and rice grains are very detailed and realistic.
- − Text spelling is incorrect, displaying 'SIIHI' instead of 'SUSHI'.
- − The flag icon is incorrect, appearing closer to the Polish flag than the Japanese flag.
- − The text is placed on a sign and the plate rather than floating at the top-center as requested.
Verdict: FLUX.1 Kontext [pro] followed the prompt instructions near-perfectly, delivering a clean, professional-grade 3D icon with correct spelling and a focused composition. Stable Diffusion 3.5 Large Turbo struggled significantly with the text rendering and the specific placement of elements, and also failed to generate the correct national 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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Follows the prompt perfectly by including all four distinct animals: golden retriever, tabby kitten, bunny, and fox kit.
- + Excellent lighting effects with soft god rays and atmosphere that feels warm and golden.
- + The fur textures are incredibly detailed and maintain a realistic, fluffy appearance.
- − One of the kittens has slightly unnatural eye placement looking at the camera compared to the others.
- − The animals are mostly sitting rather than 'tumbling' or 'chasing' as described.
Stable Diffusion 3.5 Large Turbo
- + High contrast and vibrant colors make the image pop.
- + Backlighting on the fur creates a nice rim light effect.
- − Failed prompt adherence by missing the fox kit and the bunny entirely.
- − The art style is more illustrative and smooth rather than 'hyper-photorealistic'.
- − Visible anatomical issues, such as a paw appearing to grow out of the dog's chest.
Verdict: FLUX.1 Kontext [pro] is the clear winner as it successfully rendered all four requested animals with high fidelity fur textures and beautiful atmospheric lighting. Stable Diffusion 3.5 Large Turbo failed to include the bunny and fox, and the resulting image has a more plastic, CGI look with several anatomical glitches.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Excellent typography with a clean, classic serif font.
- + Perfectly captures the minimalist vector emblem style requested.
- + Followed all text instructions including the 'Est. 1720' banner accurately.
- − Minor spelling error in the banner with 'EEST' instead of 'EST'.
Stable Diffusion 3.5 Large Turbo
- + Beautifully detailed illustration style with a vintage feel.
- + Creative integration of a coffee mug and cloche dome.
- + Atmospheric background texture adds to the retro theme.
- − Misspelled the primary name as 'Caffeé Florin' instead of 'Caffè Florian'.
- − Typography on the banner is cramped and less legible compared to Model A.
- − Cluttered composition for a 'minimalist' logo request.
Verdict: FLUX.1 Kontext [pro] creates a much more authentic minimalist logo that adheres to the requested vector emblem style, despite a small typo in the establishing date. Stable Diffusion 3.5 Large Turbo produces a more complex illustration that fails the 'minimalist' requirement and contains significant spelling errors in the main brand name.
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.”
FLUX.1 Kontext [pro]
Stable Diffusion 3.5 Large Turbo
AI Judge Analysis
FLUX.1 Kontext [pro]
- + Successfully includes most specific mission labels like 'Translunar' and 'Tranquility'.
- + Follows the NASA color palette and clean vector aesthetic well.
- + Captures the sense of a step-by-step orbital diagram.
- − Includes nonsensical elements like Saturn-style rings on both the rocket and the Moon.
- − Text labels for steps 2 and 4 are jumbled or misplaced relative to the Earth and Moon.
- − The rocket design is a generic retro rocket rather than the requested Saturn V.
Stable Diffusion 3.5 Large Turbo
- + Strong composition that feels like a professional modern infographic poster.
- + Excellent use of the 'flat-vector' style with crisp lines and limited palette.
- + Text is arranged in a clear list format that mimics an infographic structure.
- − Severe spelling errors in major headings ('Apoll.o', 'Laurch', 'TransLuna').
- − The list of steps is incomplete and does not clearly map to the visual icons.
- − Failed to provide the specific Saturn V or Lunar Module icons requested in the prompt.
Verdict: Both models struggled with the specific technical icons and accurate scientific representation. FLUX.1 Kontext [pro] followed the labeling requirements better, though it added strange planetary rings to every object; Stable Diffusion 3.5 Large Turbo produced a superior layout and artistic style but failed significantly on text legibility and following the specific sequence of six steps.
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