FLUX.2 [dev] Flash vs FLUX.2 [dev] Turbo
Head-to-head across 9 challenges
FLUX.2 [dev] Flash
25.0%
win rate
Ties
12.5%
FLUX.2 [dev] Turbo
62.5%
win rate
Challenge Results
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.2 [dev] Flash
- + Excellent prompt adherence, correctly following all positional and color instructions.
- + Realistic glass textures including subtle dust and refractions.
- + Accurate lighting and shadows coming from the left window source.
- − The plant in the background is slightly less defined compared to Model B.
FLUX.2 [dev] Turbo
- + Excellent photographic quality and sharp focus on the glass cube.
- + Very realistic wood grain on the table.
- + Clear visibility of the plant through the glass as requested.
- − The back-right vertical edge of the cube appears slightly misaligned with the top corner.
- − The blue sphere has a slightly more artificial appearance compared to Model A.
Verdict: Both FLUX.2 [dev] Flash and FLUX.2 [dev] Turbo produced nearly identical, high-quality results that followed the prompt perfectly. FLUX.2 [dev] Flash is slightly preferred for its superior rendering of the glass cube's structural integrity and the more natural integration of the blue sphere within the scene.
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.2 [dev] Flash
- + Excellent adherence to the 'imperfect framing' prompt with a tight, slightly off-center crop.
- + Very realistic skin texture and age spots on the man's hands and face.
- + Stronger bokeh effect and light reflections that enhance the cinematic feel.
- − The front wheel is missing its tire/rim at the bottom, appearing to clip into the ground.
- − The man's right hand has some minor anatomical blurring/merging with the bike frame.
FLUX.2 [dev] Turbo
- + Better overall composition that shows more of the rainy street environment and background activity.
- + More natural rain effect visible against the darker background elements.
- + The tools and parts on the ground add to the storytelling of a 'repair' scene.
- − The front wheel has a significant rendering error where the rim is severed at the bottom.
- − The man's posture looks slightly uncomfortable as if he is kneeling inside the bike frame.
Verdict: Both models captured the requested atmosphere incredibly well, particularly the 'candid' look and the motion blur of passing cars. FLUX.2 [dev] Flash produces a more intimate, high-quality portrait with superior skin textures, while FLUX.2 [dev] Turbo offers a better sense of place but suffers from more noticeable structural glitches in the bicycle's wheels.
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.2 [dev] Flash
- + Excellent symmetry in the armor engraving
- + Clean and well-defined braid structure
- + Superior leather texture and stitching on the straps
- − The torch positions are very symmetrical, feeling slightly artificial
- − Sparks are somewhat uniform in size
FLUX.2 [dev] Turbo
- + More natural and dynamic lighting from the single strong torch
- + Greater variety in spark size and bokeh depth
- + Excellent facial texture and grit
- − The leather strap buckle on the chest is physically impossible/confusing
- − Less symmetrical armor engraving compared to Model A
Verdict: Both models followed the prompt exceptionally well, capturing the beads, scars, and textures perfectly. FLUX.2 [dev] Flash achieves a cleaner, more balanced composition with better mechanical details on the buckles, while FLUX.2 [dev] Turbo has more realistic, atmospheric lighting and better spark distribution.
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.2 [dev] Flash
- + Excellent text rendering with no spelling errors.
- + Consistent handwriting style across all menu items.
- + Natural chalk texture and smears look highly realistic.
- − The title is on a single line which makes it slightly cramped near the edges.
FLUX.2 [dev] Turbo
- + Better compositional layout for the title area.
- + Realistic chalk texture and background café atmosphere.
- + Accurately followed the prompt's content requirements.
- − Contains a formatting error/repetition in the first item price ('- $ - $24').
- − The handwriting style is slightly more inconsistent between the title and the body text.
Verdict: Both models demonstrate exceptional text-rendering capabilities, but FLUX.2 [dev] Flash is the winner for its flawless execution of the text. While FLUX.2 [dev] Turbo has a slightly better layout for the header, it introduced a digital-like artifact/error in the pricing of the Risotto, whereas FLUX.2 [dev] Flash perfectly captured every word and dollar amount as 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.2 [dev] Flash
- + Excellent photorealism with smooth lighting on the capybara's fur.
- + Consistent clothing textures for both the driver and the passenger.
- + The capybara's paws are anatomically believable in their grip on the wheel.
- − The passenger appears to be in the front passenger seat rather than the requested back seat.
- − The 'TAX' sign on the roof is slightly garbled and looks like 'TAI' or 'TAX' with artifacts.
FLUX.2 [dev] Turbo
- + The taxi driver cap has a more authentic chauffeur/taxi style brim.
- + High-quality rendering of the city lights through the windshield.
- + Accurate bored expression on the passenger's face.
- − The passenger is seated in the front seat instead of the requested back seat, which changes the dynamic of the scene.
- − The capybara's paws have slightly odd, long claws that look less natural than model A.
Verdict: Both models failed to place the passenger in the back seat, instead placing her in the front passenger seat. FLUX.2 [dev] Flash is slightly preferred for its superior fur rendering and lighting continuity, though both images are remarkably similar.
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.2 [dev] Flash
- + Excellent typography and layout for the main title.
- + High-contrast, spooky atmospheric lighting around the jack-o-lantern.
- − Includes some gibberish text 'Hunk: 0369' in the details section.
- − Vignetting is a bit heavy, making the lower corners harder to read.
FLUX.2 [dev] Turbo
- + Perfect text rendering for all requested details with no extra gibberish.
- + Better overall color palette that conveys the vintage parchment feel.
- + Superior border details with clearer thorn and web integration.
- − The 'Party Invitation' text has slightly less artistic flourishing than Image A.
Verdict: Both models followed the prompt exceptionally well, producing highly polished cinematic images. FLUX.2 [dev] Turbo (Image B) is the winner because it rendered the event details perfectly without adding the nonsensical text found in FLUX.2 [dev] Flash (Image A).
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.2 [dev] Flash
- + Perfect text rendering and layout for 'JAPAN' and 'SUSHI'.
- + Clean, soft textures that perfectly match the 'cartoon scene' and 'refined textures' prompt.
- + Excellent miniature diorama feel with a simple, high-quality wooden base.
- − The sushi construction is slightly surreal, blending nigiri toppers with a maki-like interior core.
FLUX.2 [dev] Turbo
- + Highly realistic rice and fish textures with great PBR material properties.
- + More traditional and recognizable sushi (nigiri) presentation.
- + Accurate text and flag placement.
- − The wooden base has some texture tiling/clipping issues on the left edge.
- − Slightly more cluttered with additional garnishes compared to the 'minimal' request.
Verdict: Both models followed the prompt exceptionally well, adhering to the isometric perspective, text requirements, and color palette. FLUX.2 [dev] Flash is preferred as it better captures the 'cartoon scene' and 'soft refined texture' aesthetic requested, whereas FLUX.2 [dev] Turbo leans a bit more towards photo-realism with slightly less clean diorama edges.
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.2 [dev] Flash
- + Includes all four requested animals plus an extra rabbit, creating a fuller scene.
- + Excellent fur textures particularly on the fox and the golden retriever.
- + The lighting and 'dew sparkles' are clearly defined and aesthetically pleasing.
- − The extra animal (second rabbit/bunny head) makes the anatomy in the center look slightly crowded and confusing.
- − The cat's posture is a bit stiff compared to Model B.
FLUX.2 [dev] Turbo
- + Perfect adherence to the count and types of animals requested (puppy, kitten, bunny, fox).
- + Dynamic and playful posing that better captures the 'tumbling' and 'chasing' part of the prompt.
- + Very high visual clarity and consistent lighting across the different fur types.
- − Slightly fewer butterflies than Model A, though still sufficient for the prompt.
- − The kitten's tail looks slightly elongated.
Verdict: Both models produced high-quality, charming images that captured the golden hour lighting and dew sparkles. FLUX.2 [dev] Turbo is the winner because it followed the specific animal count accurately and captured more dynamic, 'tumbling' poses, whereas FLUX.2 [dev] Flash added an extra rabbit that made the center composition look a bit cluttered.
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.2 [dev] Flash
- + Clean, legible typography for the main headers and crew members.
- + Very high-quality vector illustrations of the Lunar Module and Saturn V.
- + Good adherence to the requested NASA-inspired color palette.
- − Literal interpretation of the prompt led to 'Satauuri Iccon' appearing as text rather than just an icon.
- − Includes duplicate labels for 'Descent' and 'Landing' creating a cluttered layout.
- − Minor spelling errors in small supporting text (e.g., 'Tranquility' vs 'Tranquility').
FLUX.2 [dev] Turbo
- + Includes a well-rendered NASA logo that fits the prompt's theme.
- + Professional layout with better use of negative space.
- + Accurate spelling of 'Tranquility' and the crew names.
- − Muddied the requested workflow by duplicating 'Translunar' and the Moon icons.
- − Also included technical prompt text ('Saturn V icon') as a text label.
- − The trajectory paths are a bit confusing and overlap inconsistently.
Verdict: Both models struggled with the 'negative constraint' of not including prompt instructions as literal text labels (both included variations of 'Saturn Icon'). Flux.1 [dev] Flash produces slightly more detailed individual illustrations, but Flux.1 [dev] Turbo has a cleaner, more professional infographic layout and better spelling. Turbo is the preferred choice for a cohesive poster design despite some redundancy in the icons.
FLUX.2 [dev] Flash
Fast distilled version of Black Forest Labs' FLUX.2 [dev] optimized for speed and cost efficiency.
FLUX.2 [dev] Turbo
Distilled version of Black Forest Labs' FLUX.2 [dev] outperforming it at a cheaper price. Developed by fal.ai.