Black Forest Labs' 12 billion parameter distilled image generation model optimized for speed, capable of generating high-quality images in just 4 inference steps
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 [schnell]
18.7
arena score
#48 of 62 in Text-to-Image
Skill signature
· Text-to-Image
FLUX.2 [klein] 9B
20.6
arena score
#13 of 32 in Image Editing
Vote tally
Where the votes landed
FLUX.1 [schnell]
0%
win rate
Ties
0%
FLUX.2 [klein] 9B
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Vibrant colors and high resolution
- + Excellent refractive and reflective details in the glass
- + Follows most spatial relationships correctly
- − Added an extra blue sphere on top of the book which was not requested
- − The central sphere appears to be floating without visible support
FLUX.2 [klein] 9B
- + Accurately depicts all requested elements without adding unwanted objects
- + Realistic soft window lighting and shadows on the table
- + The sphere appears more naturally sized ('small') relative to the cube
- − The glass cube is curved/rounded and looks more like a vase or jar
- − The plant behind the cube is quite blurry and less integrated
Verdict: FLUX.1 [schnell] creates a more visually striking and polished image, but fails the text-to-image challenge by adding a second blue sphere on top of the book. FLUX.2 [klein] 9B follows the prompt instructions perfectly, placing exactly one sphere inside the cube and one book on top, making it the more accurate model despite the slightly rounded edges of 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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Strong bokeh and cinematic lighting in the background
- + Vibrant color on the red bicycle
- + Good wet pavement reflections
- − The man's hands are anatomically incorrect and fused with the handlebars
- − Fails to show actual repair work as requested
- − Cars in the background lack the requested motion blur
FLUX.2 [klein] 9B
- + Excellent adherence to the 'repairing' action with visible tools
- + Very realistic skin texture and facial features
- + Successfully captures the specific requested framing and motion blur on the background car
- − Minor floating brake cable near the front fork
- − The bicycle's perspective is slightly warped near the front tire
Verdict: FLUX.2 [klein] 9B is the clear winner as it accurately captures the 'repairing' aspect of the prompt and includes tools, whereas FLUX.1 [schnell] shows a man simply leaning on a bicycle. FLUX.2 [klein] 9B also delivers a much more realistic human figure with natural skin texture and better adherence to 'motion blur' and 'candid' attributes.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Intense, high-contrast lighting that creates a dramatic mood
- + Extremely sharp facial skin textures and detailed iris rendering
- + Very close framing that emphasizes the 'close portrait' instruction
- − The armor is mostly cropped out, missing the 'ornate engraved plate' request
- − The braids are messy and poorly defined compared to Model B
- − Over-sharpening artifacts on the forehead wrinkles
FLUX.2 [klein] 9B
- + Excellent adherence to all prompt details including braids with beads and engraved armor
- + Superior material rendering on leather straps, mail, and scratched plate metal
- + Beautiful bokeh effect with floating sparks and torchlight
- − The facial features are slightly soften compared to the hyper-detailed armor
- − The eyes, while good, lack the piercing intensity seen in Model A
Verdict: Model B (FLUX.2 [klein] 9B) is the clear winner as it successfully incorporates every specific element of the prompt, including the complex engraved armor and the braided hair with beads, whereas FLUX.1 [schnell] cropped out much of the costume. Model B also provides a much better sense of environment with the bokeh sparks and visible torches, creating a more cohesive story for the character.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Clean, professional minimalist aesthetic
- + Good hierarchical structure in text layout
- + Accurate representation of the requested food sections
- − Internal food photography is slightly repetitive and lacks variety
- − Text rendering for body copy consists of illegible characters
- − A bit too much empty white space on the right side
FLUX.2 [klein] 9B
- + Dynamic and colorful grid of high-quality food photography
- + Excellent use of bold typography and vibrant orange/blue accents
- + Well-organized two-column layout that feels like a real menu
- − Major spelling error in the main header text
- − Section headings are redundant (Appetizers listed twice)
- − Includes bizarre pricing characters like '$G' and '$0'
Verdict: FLUX.1 [schnell] creates a more believable professional template with a clean minimalist vibe, although it struggles with body text legibility. FLUX.2 [klein] 9B is much more vibrant and creative with its use of color and food photography, but it suffers from poor text logic, including repeating the same header twice and misspelling the largest word on the page.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Excellent photorealistic texture on the meat patty and bun
- + Creatively interprets the 'exploded' request with surrounding ingredients
- − Significant text errors including 'AGIC BURGER' and a duplicated price of '€699'
- − The main burger is mostly assembled rather than exploded
FLUX.2 [klein] 9B
- + Perfect text rendering for all requested phrases including the currency symbol
- + Effective 'fiery' glow effect on the typography as requested
- + Strong sense of motion with the radial sparks and tilted composition
- − Failed to 'explode' the burger, showing it almost entirely assembled
- − The floating sauce drops look a bit like plastic or solid objects
Verdict: FLUX.2 [klein] 9B is the clear winner as it precisely followed the complex text requirements, including the fiery glow effect and correct pricing. While FLUX.1 [schnell] produced a highly detailed burger, it failed significantly on the typography, misspelling the title and creating confusing price labels.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Clean layout that is easy to read.
- + Accurately represents a cafe setting environment.
- − Numerous spelling errors including 'Pril', 'Taffle', 'Mushmnctionm', and 'Octtoopus'.
- − Fails the request for elegant cursive handwriting for the title.
- − The chalk texture is too clean, looking more like a digital marker than real chalk.
FLUX.2 [klein] 9B
- + Excellent spelling accuracy for all requested menu items.
- + Beautifully rendered chalk texture with realistic smudges and strokes.
- + Perfectly follows the stylistic request for elegant cursive in the title.
- − Includes a minor typo in the footer text 'fress' instead of 'fresh'.
- − The layout is a bit tight at the bottom.
Verdict: FLUX.2 [klein] 9B followed every instruction of the prompt, including complex spelling and specific stylistic requests for cursive and chalk texture. FLUX.1 [schnell] failed significantly on text rendering, producing several nonsensical words and ignoring the cursive requirement.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.1 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Perfectly follows the specific role-reversal instruction of the horse riding the astronaut.
- + Cinematic lighting and high-quality textures on the space suit.
- + Creative interpretation of the horse's anatomy to fit the surreal prompt.
FLUX.2 [klein] 9B
- + Excellent background detail with various celestial bodies.
- + High resolution and clear rendering of the astronaut's face and gear.
- − Completely failed the negative constraint/instruction for the horse to be on top.
- − Nonsensical text at the bottom of the image.
- − Typical, non-surreal interpretation of the prompt.
Verdict: FLUX.1 [schnell] followed the complex prompt instruction to place the horse on top of the astronaut, creating a truly surreal image. FLUX.2 [klein] 9B ignored the specific 'horse on top' instruction, providing a standard astronaut-on-horse image with unwanted gibberish text.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Excellent fur texture and lighting on the capybara
- + High cinematic quality with professional-looking bokeh
- + Accurate interpretation of 'bored expression' on the passenger
- − One paw is missing from the steering wheel
- − Anatomically confused front leg placement
- − Text on the visor is nonsensical gibberish
FLUX.2 [klein] 9B
- + Perfect adherence to the 'both front paws on the steering wheel' instruction
- + Realistic cap-style for a taxi driver
- + Better spatial logic for a vehicle interior
- − Passenger appears to be sitting in the front passenger seat rather than the back as requested
- − Text on the hat is slightly garbled ('NEW TALA')
- − Capybara scale feels slightly off compared to the car seat
Verdict: While FLUX.1 [schnell] has superior cinematic lighting and textures, it failed the specific pose instruction regarding the paws on the steering wheel. FLUX.2 [klein] 9B followed the pose instructions perfectly and provided a more readable composition, even though it placed the passenger in the front seat instead of the back. Overall, FLUX.2 [klein] 9B is the winner for its better handling of the capybara's interaction with the car controls.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Successfully included all elements like bats, trees, and jack-o-lantern
- + The central jack-o-lantern has a nice internal glow effect
- − Significant text errors including typos and repeated lines of nonsense
- − The layout is cluttered with unnecessary repeat banners
- − Graphic design quality looks dated and less cohesive
FLUX.2 [klein] 9B
- + Perfect text rendering for all requested copy including complex details
- + Much better adherence to the 'webs and thorns' border request
- + Vibrant cinematic lighting and a more professional illustration style
- − The parchment aging on the edges is a bit uniform
- − The moon is partially obscured by the tree in a way that slightly affects balance
Verdict: FLUX.2 [klein] 9B followed every instruction perfectly, including the specific text strings and the complex web/thorn border. In contrast, FLUX.1 [schnell] struggled with the text, producing several typos and redundant lines that made the invitation unusable. FLUX.2 [klein] 9B also provided a much more polished, cinematic aesthetic that fits the 'vintage gothic' theme.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Excellent clean 3D render aesthetic with smooth lighting
- + Very faithful to the requested isometric diorama base
- + Correct Japanese flag icon
- − Missed the word 'SUSHI' in the text overlay
- − The salmon texture on the sushi looks slightly stretched or distorted
FLUX.2 [klein] 9B
- + Successfully included all requested text ('JAPAN' and 'SUSHI')
- + Includes a variety of sushi items on the plate
- + Great clay-like cartoon textures that match the 'miniature' prompt
- − Failed to render the correct flag of Japan, showing a tricolor flag instead
- − The wood grain on the base has some tiling artifacts
Verdict: Both models followed the isometric diorama style well. FLUX.1 [schnell] produced a cleaner, more professional-looking render with the correct flag, but failed to include the word 'SUSHI'. FLUX.2 [klein] 9B captured all text elements and provided more variety in the composition, but suffered from a significant error regarding the flag of Japan.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Excellent fur texture rendering
- + Cohesive warm lighting effect
- − Failed to include the four distinct species requested
- − The third animal is an anatomical hybrid of a cat and bunny without clear identity
- − Lacks the requested god rays
FLUX.2 [klein] 9B
- + Captured the dynamic 'playfully chasing' action much better
- + Successfully included god rays and dew sparkles as per the prompt
- + Superior composition with more interaction between animals
- − Failed to include the bunny
- − The kitten has an oddly wide-eyed, startled expression
Verdict: While both models failed to generate all four distinct animal species (both missing the bunny), FLUX.2 [klein] 9B followed the environmental prompts for god rays and dew sparkles more effectively. Additionally, FLUX.2 [klein] 9B captured the energetic, playful atmosphere of the scene, whereas FLUX.1 [schnell] produced a more static, slightly repetitive composition with anatomical ambiguities.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Strong minimalist vector aesthetic
- + Good use of warm brown and cream tones
- − Significant text errors ('Framilan' instead of 'Florian', '7720' instead of '1720')
- − Missing the 'steam' element requested in the prompt
FLUX.2 [klein] 9B
- + Perfect text accuracy and spelling
- + Includes all requested elements including the cloche with steam and the specific banner
- + Excellent paper-like subtle texture on the background
- − The 'steam' graphic is slightly cartoonish compared to the rest of the logo
Verdict: FLUX.2 [klein] 9B followed the prompt instructions perfectly, including the correct spelling of 'Caffè Florian' and the accurate date. FLUX.1 [schnell] failed significantly on the typography, hallucinating both the name and the date, and missed the steam element entirely.
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 [schnell]
FLUX.2 [klein] 9B
AI Judge Analysis
FLUX.1 [schnell]
- + Successfully captures a minimalistic, vector aesthetic consistent with modern infographics.
- + Uses the requested NASA-inspired color palette effectively.
- + Includes a clear landing representation at the bottom.
- − Text is largely illegible gibberish.
- − The composition is confusing and does not clearly follow a linear 6-step progression.
FLUX.2 [klein] 9B
- + Shows better understanding of the step-by-step nature of the request.
- + Includes recognizable labels for the mission phases despite some spelling errors.
- + The icons for the rocket and Lunar Module are cleaner and more accurate to the prompt.
- − Significant spelling errors in prominent text (e.g., 'AOILLO 11', 'LANDINING').
- − The flow of the infographic is a bit cluttered compared to a professional vector poster.
Verdict: While FLUX.1 [schnell] captures a more polished 'vector art' style, it fails to present the information in a logical or readable way. FLUX.2 [klein] 9B followed the instructional nature of the prompt much better, providing distinct sections for the requested mission steps, making it the more functional infographic despite spelling errors.
Next steps
Explore each model
Black Forest Labs' distilled 9 billion parameter image generation model with sub-second inference and multi-reference support