FP8 quantized variant of Black Forest Labs' FLUX.1 [schnell] model, offering ~2x faster inference with reduced precision while maintaining high-quality image generation in 4 steps
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 [schnell] FP8
#47 of 62 in Text-to-Image
FLUX.2 [dev] Turbo
#3 of 62 in Text-to-Image
Where the votes landed
FLUX.1 [schnell] FP8
0%
win rate
Ties
0%
FLUX.2 [dev] Turbo
0%
win rate
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Features very clean, sharp rendering of the glass and sphere
- + Strong lighting effects with nice surface reflections
- + Vibrant colors that make the subjects pop
- − The glass object is a tall rectangular prism rather than a cube
- − The sphere appears to be floating on an internal shelf rather than sitting on the bottom of the cube
- − The plant is mostly beside the glass rather than behind it
FLUX.2 [dev] Turbo
- + Accurately depicts the object as a cube
- + Excellent prompt adherence regarding the plant being partially visible through the glass
- + Highly realistic textures on the wooden table and the book cover
- − Lighting is somewhat muted compared to the prompt's request for window light
- − Slightly less 'polished' aesthetic compared to the first image
Verdict: While both models followed the prompt well, FLUX.2 [dev] Turbo succeeded in more complex spatial requirements, such as correctly rendering a cube shape and showing the plant through the glass. FLUX.1 [schnell] FP8 produced a more vibrant image, but failed on the specific geometry of a cube and the placement of the plant.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Excellent shallow depth of field and bokeh effect
- + Atmospheric use of light and color reflections
- + Clean rendering of the bicycle frame
- − Cars are stationary rather than having the requested motion blur
- − The man's hands are fused incorrectly into the middle of the handlebars
- − Lacks visible falling rain as requested
FLUX.2 [dev] Turbo
- + Accurately depicts light rain and motion blur on passing cars
- + Realistic skin texture and age-appropriate details on the subject
- + Detailed 'repair' scene with tools on the ground and hands interacting with the bike
- − The bike's front wheel is missing its bottom half where it meets the puddle
- − The kickstand and frame geometry are slightly warped
- − The background is a bit busier, which slightly offsets the focus from the subject
Verdict: FLUX.2 [dev] Turbo followed the complex instructions much better, delivering visible rain, motion blur, and a character actually engaged in a repair. While FLUX.1 [schnell] FP8 produced a more traditionally beautiful bokeh, it failed on the specific technical prompt requirements like motion blur and rain, and had significant structural errors in the hands and handlebars.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Intense, high-contrast lighting that creates a dramatic mood.
- + Extreme levels of skin texture and facial detail.
- + Clear depiction of shallow depth of field with soft bokeh.
- − Missed the request for 'braided hair with small beads' almost entirely.
- − The armor is less prominent and detailed than the face.
- − Lacks the specific 'scars' requested in the prompt.
FLUX.2 [dev] Turbo
- + Excellent adherence to all prompt details including braids, beads, and visible scars.
- + Superior rendering of 'ornate engraved plate armor' and associated leather straps.
- + Captures the atmosphere perfectly with a visible torch and flying sparks.
- − The lighting is a bit flat across the face compared to Model A.
- − Some of the sparks look like simple digital streaks rather than physical embers.
Verdict: FLUX.2 [dev] Turbo is the clear winner as it followed every instruction in the prompt, including the specific requests for hair beads, scars, and ornate engravings. While FLUX.1 [schnell] FP8 produced a striking portrait, it neglected the majority of the thematic details like the braids and the battle-worn features.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Includes all requested text sections (Appetizers, Pizza, Mains)
- + Mimics a multi-page physical menu layout well
- − Text is largely gibberish or misspelled ('ORCETERS', 'APPTIZERS', 'SECCER')
- − The grid layout of images is inconsistent in size and alignment
- − Overall image quality is somewhat blurry
FLUX.2 [dev] Turbo
- + Excellent photographic quality and color vibrance
- + Stronger adherence to 'modern minimalist' and 'vibrant accents' with the color bar at the bottom
- + Clean, bold sans-serif typography is highly legible
- − Sectioning is confusing, putting clearly pizza-themed photos under 'Appetizers'
- − Text content is nonsensical and includes strange pricing like $300
Verdict: FLUX.2 [dev] Turbo followed the aesthetic cues of the prompt much better, delivering a clean, modern design with a professional color palette. While FLUX.1 [schnell] FP8 attempted to include more sections, its poor spelling and lower visual clarity make it less usable than FLUX.2 [dev] Turbo's polished, high-contrast output.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Features a bold and clear composition
- + The burger itself is vibrant and well-rendered
- − Massive spelling errors: 'LIIMITED', 'LIMIED', 'NEEY'
- − Incorrect price displayed and poor placement of the starburst
- − Failed to apply requested fiery, glowing effects to the text
FLUX.2 [dev] Turbo
- + Perfect text rendering for all requested phrases including the price
- + Successfully applied the 'fiery, glowing effect' to all typography
- + Better representation of an 'exploded' burger with components suspended in mid-air
- − The lettuce is somewhat clumped together rather than individually exploding
- − The background texture is slightly noisier than Image A
Verdict: FLUX.2 [dev] Turbo is the clear winner as it accurately followed every instruction, including the complex text rendering and fiery styling. FLUX.1 [schnell] FP8 failed significantly on the typography, producing several typos and completely ignoring the requested visual style for the text.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Clean layout and frame
- + Good lighting and focus
- − Significant spelling errors and repeated text lines
- − The text looks more like a digital marker font than realistic chalk
- − Failed to follow the cursive request for the title
FLUX.2 [dev] Turbo
- + Excellent adherence to the 'realistic chalk' texture and variations
- + Perfect spelling and completion of the menu items as requested
- + Successfully used elegant cursive for the title
- − Slightly more cluttered background compared to Image A
Verdict: FLUX.2 [dev] Turbo significantly outperformed FLUX.1 [schnell] FP8 by adhering to every specific detail of the prompt, including the cursive title and accurate spelling of complex menu items. While Image A suffered from severe text artifacts and repeated nonsense words, Image B produced a highly realistic, legible, and aesthetically pleasing chalkboard that looks authentically hand-drawn.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Attempts a surreal interpretation close to the prompt's structural requirement
- + Features high-quality lighting and cinematic depth of field
- − Anatomy is nonsensical with multiple horse heads and distorted bodies
- − Fails to clearly depict an astronaut under the horse, showing only abstract machinery
FLUX.2 [dev] Turbo
- + Excellent visual clarity and technical rendering of textures
- + Beautiful cinematic background with high detail in the astronaut suit and horse
- − Directly fails the core prompt instruction of 'horse on top, not vice versa'
- − Conventional interpretation lacks the requested surrealism
Verdict: This is a case of choosing between technical execution and instruction following. FLUX.1 [schnell] FP8 attempted the difficult 'horse on top' request but resulted in a chaotic, anatomical mess, while FLUX.2 [dev] Turbo created a stunning image that completely ignored the inversion instruction. FLUX.2 is the better image visually, but neither model successfully realized the specific surreal concept requested.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Successfully captures both front paws on the steering wheel.
- + The capybara's expression is very calm and focused.
- + The framing through the window creates an intimate interior feel.
- − The woman is holding two phones simultaneously, which is illogical.
- − The capybara's hat looks more like a plastic toy than a driver's cap.
- − The proportions of the car's interior feel slightly cramped and artificial.
FLUX.2 [dev] Turbo
- + Higher realism in the taxi's exterior and interior details.
- + Captures the 'bored' expression of the businesswoman perfectly.
- + The capybara's clothing and hat look authentic and professional.
- − The capybara's paws look somewhat mangled and unnatural on the steering wheel.
- − The framing places the passenger in the front seat instead of the 'back seat' as requested.
- − Minor spelling error on the roof sign ('TAX').
Verdict: Both models struggled with certain specific instructions: FLUX.1 [schnell] FP8 followed the 'back seat' instruction but failed on basic logic by giving the woman two phones. FLUX.2 [dev] Turbo produced a far more photorealistic and atmospheric image with more believable clothing and textures, but it placed the passenger in the front seat and had issues with the capybara's paws.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Atmospheric cinematic lighting on the pumpkin
- + Good use of negative space for a moody effect
- − Numerous spelling errors in the text
- − The layout of the event details is messy and confusing
- − Fails to include the thorns and webs in the border explicitly
FLUX.2 [dev] Turbo
- + Perfect text rendering for both titles and event details
- + Highly detailed border featuring thorns and spider webs as requested
- + Excellent vintage gothic aesthetic with a cohesive composition
- − The lighting on the pumpkin is slightly flatter than the cinematic approach in image A
Verdict: FLUX.2 [dev] Turbo significantly outperforms FLUX.1 [schnell] FP8 by adhering perfectly to all textual requirements and visual details, including the thorns and webs in the border. FLUX.1 [schnell] FP8 suffers from significant 'AI gibberish' spelling across the entire bottom half of the image.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Excellent soft refined textures and miniature 3D cartoon style
- + Follows the isometric perspective very well
- + Clean aesthetic with a professional-looking diorama base
- − Failed the text prompt, repeating 'JAPAN' incorrectly as 'JA-AN'
- − Missing the word 'SUSHI' entirely
FLUX.2 [dev] Turbo
- + Perfect text rendering of 'JAPAN' and 'SUSHI'
- + High-quality PBR materials, especially on the salmon and seaweed
- + Accurate inclusion of the flag icon and wood texture on the base
- − The lighting is slightly flatter compared to Model A
- − The shadows on the background appear a bit grainy
Verdict: Model B (FLUX.2 [dev] Turbo) significantly outperformed Model A in prompt adherence, correctly rendering all requested text and icons. While Model A (FLUX.1 [schnell] FP8) captured the 3D cartoon style beautifully, its failure to generate the correct text makes Model B the clear winner.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Captures the golden sunrise and god rays very effectively
- + Creates a high-contrast, dreamlike aesthetic that matches the 'masterpiece' keyword
- + Fur texture appears very soft and voluminous
- − Failed to include the baby bunny, instead generating extra kittens/unidentifiable kits
- − The animals are mostly static and posing rather than 'tumbling'
- − Anatomical issues with the animals merging into one another
FLUX.2 [dev] Turbo
- + Correctly included all requested animals: puppy, tabby kitten, bunny, and fox kit
- + Captured the 'tumbling' and 'playful chasing' action much better than Model A
- + Excellent rendering of dew sparkles and realistic fur detail
- − The fox kit has a slightly awkward extra-long tail and paw placement
- − Lighting is a bit flatter compared to the intense god rays requested
Verdict: FLUX.2 [dev] Turbo provided a much better adherence to the prompt by including the specific animal variety requested and capturing a sense of movement. FLUX.1 [schnell] FP8 failed to generate the bunny and produced a more cluttered, static composition, although it had stronger lighting effects.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Successfully captures the warm brown and cream color scheme
- + Clean vector-style execution
- − Major spelling errors in the brand name ('Afe Flamilan')
- − Incorrect iconography, using a dome or cup lid instead of a cloche
- − Messy typography arrangement within the banner
FLUX.2 [dev] Turbo
- + Perfect text rendering for both name and date
- + Accurate depiction of a cloche dome with steam as requested
- + Excellent use of subtle texture for a vintage feel
- − The steam trails are slightly asymmetrical
Verdict: FLUX.2 [dev] Turbo followed all instructions perfectly, including the specific iconography and correct spelling of 'Caffè Florian'. FLUX.1 [schnell] FP8 failed significantly on the text, generating nonsensical characters and missing the core cloche element, instead providing a generic building or lid silhouette.
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.”
AI Judge Analysis
FLUX.1 [schnell] FP8
- + Strong minimalist vector aesthetic
- + Excellent adherence to the color palette constraints
- + Clean layout with a structured timeline feel
- − Nonsense text throughout the image
- − Iconography is overly abstract and lacks representational detail
- − Missed several specific steps including the Saturn V and Descent icons
FLUX.2 [dev] Turbo
- + Excellent text rendering with accurate labels and names
- + High level of detail in icons while maintaining a vector look
- + Includes all requested steps and adds relevant extra details like the crew
- − The layout is a bit cluttered compared to a professional infographic
- − Deviates slightly from the 'muted' palette by including bright gold/yellow on the modules
- − The 'Translunar' label is repeated twice in the middle of the composition
Verdict: FLUX.2 [dev] Turbo is the clear winner as it successfully rendered most of the requested text and steps, creating a highly informative and visually appealing infographic. While FLUX.1 [schnell] FP8 captured the minimalist vector style better, its failure to generate readable text or accurate icons for the mission steps makes it less useful for the specific prompt.
Explore each model
Distilled version of Black Forest Labs' FLUX.2 [dev] outperforming it at a cheaper price. Developed by fal.ai.