Black Forest Labs' distilled 9 billion parameter image generation model with sub-second inference and multi-reference support
Settled by community votes across 15 shared challenges, with an AI judge weighing in on each.
FLUX.2 [klein] 9B
#13 of 32 in Image Editing
OmniGen v2
#57 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [klein] 9B
0%
win rate
Ties
0%
OmniGen v2
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.2 [klein] 9B
- + Excellent photographic texture on the wooden table and book edges.
- + Realistic glass refraction of the plant and background.
- + Excellent adherence to spatial prompts including the plant behind the cube.
- − The glass cube is filled with water or has a double-walled vessel appearance rather than being a simple hollow cube.
- − The light source is a bit diffuse despite the window placement.
OmniGen v2
- + Perfectly depicts a clean, geometric hollow glass cube.
- + Vibrant colors on the sphere and book.
- + Good lighting and shadows consistent with a left-side source.
- − The plant is significantly further in the background, making it less 'partially visible through the glass' as requested.
- − The sphere appears to be floating inside the cube rather than resting on the bottom.
Verdict: FLUX.2 [klein] 9B produces a more photorealistic scene with impressive attention to how the glass refracts the plant behind it, though the 'cube' looks more like a heavy glass vase. OmniGen v2 provides a much cleaner interpretation of the requested objects, though the plant is moved too far back to interact with the glass transparency effectively. overall, FLUX.2 feels more like a real photograph while OmniGen v2 feels like a high-quality 3D render.
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 [klein] 9B
- + Excellent skin texture and facial detail
- + Realistic, messy environment with accurate tools and bike components
- + Effective shallow depth of field and street reflections
- − Passing car lacks significant motion blur as requested
OmniGen v2
- + Strong reflections on the wet pavement
- + The subject is clearly an elderly Japanese man
- − Fails to show the man 'repairing' the bike; he is just holding it
- − Heavy AI artifacts on the bike spokes and pedals
- − The background trees and rain look stylized rather than realistic
Verdict: FLUX.2 [klein] 9B followed the prompt much more successfully, depicting a realistic repair scene with high tactile detail and authentic street atmosphere. OmniGen v2 failed the 'repairing' action and suffered from significant anatomical and structural glitches, particularly in the bicycle's design.
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 [klein] 9B
- + Excellent adherence to technical details like leather grain and chainmail textures
- + Perfect execution of the braided hair with beads requirement
- + Accurate portrayal of a battle-worn look with realistic scars and dirt
- − The lighting on the armor is a bit flat compared to the intensity of the torches
OmniGen v2
- + Natural look to the bokeh sparks and warm firelight glow
- + Pleasant composition with a dynamic angle
- − Missed the 'battle-worn' requirement as the character looks very clean and polished
- − Poor texture on the leather and armor compared to the prompt's request for high detail
- − Anatomical oddity with the braid beginning behind the ear while also appearing as a crown braid
Verdict: FLUX.2 [klein] 9B followed every detail of the prompt, specifically excelling at the battle-worn skin textures and the intricate detail of the leather and cloth layers. OmniGen v2 failed to capture the 'battle-worn' aesthetic and produced a much softer, less detailed image that disregarded several of the specific texture requests.
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.2 [klein] 9B
- + Excellent photographic quality for the food items
- + Clean, professional layout that closely mimics a real menu
- + Clearly defined sections for Appetizers, Pizza, and Mains
- − Text is largely gibberish or contains typos like 'Resteuranbreteina'
- − The header text flows into other letters uncomfortably
OmniGen v2
- + Effective use of vibrant color blocks as accents
- + Good visual variety in the types of food displayed
- + Strong adherence to the 'grid' layout requested
- − Significant spelling errors throughout ('APPTETIZES', 'PIZZZZAN')
- − The body text is reduced to illegible scribble lines
- − The composition feels slightly cluttered compared to a minimalist design
Verdict: FLUX.2 [klein] 9B produces a much more realistic and professional-looking menu with high-quality food photography that fits the 'casual dining' vibe perfectly. While OmniGen v2 has creative use of color accents, its poor text rendering and repetitive spelling errors make it less usable than FLUX.2.
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.2 [klein] 9B
- + Excellent photorealistic texture on the meat and bun
- + Dynamic 'exploded' composition with suspended sauce droplets
- + Perfect text rendering of all elements including the Euro symbol
- − The 'exploded' effect is slightly subtle, as the components are still mostly stacked
OmniGen v2
- + Strong graphic design layout that clearly resembles a fast-food ad
- + Clean and readable title text
- − Failed the 'exploded burger' requirement; the burger is fully assembled
- − Text is cut off ('LIMITED TIM...') and missing the Euro symbol
- − Image quality looks more like a 3D render/illustration than photorealistic
Verdict: FLUX.2 [klein] 9B followed every instruction, delivering a photorealistic image with accurate text and the requested 'exploded' mid-air effect. OmniGen v2 failed to explode the burger, cut off secondary text, and produced a more cartoonish result that lacked the professional polish of its competitor.
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 [klein] 9B
- + Excellent text rendering with nearly perfect spelling of all complex menu items.
- + Highly realistic chalk texture including board smudges and varying stroke opacity.
- + Successful execution of elegant cursive for the title as requested.
- − Minor spelling error in the footer text 'fress' instead of 'fresh'.
- − The year '2026' is slightly less consistent in style than the letters.
OmniGen v2
- + Natural variation in letter sizing and slant creates a casual feel.
- + Correctly positioned the frame and overall chalkboard layout.
- − Significant spelling errors throughout the menu items (e.g., 'SPECALS', 'Musonhom', 'Lemont').
- − The handwriting is messy and illegible in several places, overlapping other words.
- − Failed to maintain the 'elegant cursive' style requested for the title.
Verdict: FLUX.2 [klein] 9B followed the prompt with high precision, accurately rendering long text strings with a sophisticated chalk aesthetic. In contrast, OmniGen v2 struggled with spelling and legibility, producing garbled words and failing to deliver the specific cursive header requested.
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.2 [klein] 9B
- + Excellent cinematic rendering with complex galactic background.
- + High level of anatomical detail on the horse and suit textures.
- + Effective lighting and depth within the space scene.
- − Failed the specific spatial instruction (the astronaut is on top, not the horse).
- − Includes distracting gibberish text at the bottom.
OmniGen v2
- + Clean, high-contrast imagery.
- + Good interpretation of the space background with simpler elements.
- + Clear focus on the subject matter.
- − Failed the primary prompt instruction (the horse should be on top).
- − Low creative risk-taking, resulting in a generic 'astronaut on a horse' image.
- − Anatomical issues with the horse's back legs.
Verdict: Both FLUX.2 [klein] 9B and OmniGen v2 failed the specific negative constraint to place the horse on top of the astronaut, providing the standard 'astronaut riding a horse' trope instead. FLUX.2 [klein] 9B is the preferred image as it significantly outperforms OmniGen v2 in terms of cinematic detail, atmospheric depth, and textural realism, despite the inclusion of some AI-generated text artifacts.
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 [klein] 9B
- + Excellent adherence to the perspective, correctly placing the capybara in the driver's seat and the passenger in the back.
- + Photorealistic lighting and textures, especially on the capybara's fur and the car interior.
- + Successfully captures the specific requested expression of a bored businesswoman looking at her phone.
- − The text on the hat is distorted ('NEW TALA').
- − The capybara's front paws on the steering wheel look slightly anthropomorphic and hairy in a way that is a bit uncanny.
OmniGen v2
- + High image sharpness and vibrant colors.
- + CleanRendering of the dark jacket and cap on the capybara.
- − Failed the spatial layout significantly by putting the capybara in the passenger seat and having human hands driving the car.
- − The businesswoman is in the front seat instead of the back seat as requested.
- − The capybara's head is floating awkwardly over the seat back without a body connecting to the steering wheel.
Verdict: FLUX.2 [klein] 9B is the clear winner as it correctly followed the complex spatial instructions of having the capybara drive and the passenger sit in the back. OmniGen v2 failed the prompt by having the human drive the car and placing the capybara in the passenger seat, along with several anatomical errors regarding how the capybara is positioned.
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 [klein] 9B
- + Exceptional text rendering with perfect spelling of all requested details.
- + Superior atmospheric lighting and high-quality vintage aesthetic.
- + Excellently crafted border featuring the requested thorns and webs.
- − The parchment edges are slightly cut off by the square frame.
OmniGen v2
- + Successfully includes the requested elements like bats, trees, and jack-o-lantern.
- + The parchment-on-black background framing is visually clean.
- − Serious text issues including misspellings, repetition, and illegible pseudo-script on the banner.
- − The illustration style is more cartoonish and less 'cinematic' than requested.
- − Failed to correctly render the specific location: 'THE ARCAS, NYC' instead of 'The Arches'.
Verdict: FLUX.2 [klein] 9B followed every instruction perfectly, including complex text requirements and specific stylistic requests like 'cinematic lighting'. OmniGen v2 struggled significantly with text legibility and accuracy, producing multiple typos and failing to capture the 'polished' vintage look requested.
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 [klein] 9B
- + Excellent typography with clean, bold black text
- + High-quality soft lighting and refined wooden textures
- + Accurate 45-degree isometric perspective
- − The flag icon is incorrect, resembling the flag of Yemen or the German Empire rather than Japan
OmniGen v2
- + Bright, appealing 3D cartoon aesthetic
- + Correct text content with nice drop-shadow styling
- + Vibrant colors that pop against the background
- − The flag icon is incorrect, showing a red/yellow/blue striping
- − Minor sushi anatomy issues where the salmon tail looks stylized like a shrimp
Verdict: Both models followed the isometric and text instructions well. FLUX.2 [klein] 9B produced a more professional, 'refined' look with realistic wood grain and superior plate composition, whereas OmniGen v2 went for a bolder, more saturated cartoon style. FLUX.2 is the winner due to more consistent rendering of the 3D diorama elements and cleaner text integration.
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 [klein] 9B
- + Excellent photographic realism with detailed fur textures and realistic anatomy.
- + Successfully captures complex lighting with god rays and soft backlighting.
- + Includes three of the four requested animals with dynamic, playful posing.
- − Missed the baby bunny requested in the prompt.
- − The kitten's paws look slightly distorted during the tumbling action.
OmniGen v2
- + Bright, cheerful colors that match the wholesome vibe.
- + Good centered composition for a greeting card style.
- − Failed to follow the hyper-photorealistic style, resulting in a 3D cartoon/CGI look.
- − Missed the baby bunny and the fox kit appears more like a generic cartoon rabbit-fox hybrid.
- − Lacks the fine detail and texture quality seen in a masterpiece 8K image.
Verdict: FLUX.2 [klein] 9B followed the stylistic requirements much more effectively, producing a high-fidelity image that looks photographic and professional despite missing the bunny. OmniGen v2 failed the primary requirement for a photorealistic scene, instead generating a cartoonish image with simplified textures and anatomy. FLUX.2 clearly wins on visual quality, lighting, and detail.
Studio Ghibli Anime Style
Editing“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellently captures the requested watercolor/hand-painted texture common in Ghibli backgrounds.
- + Maintains the specific lighting and pastel color palette requested.
- + Preserves the composition and clothing patterns of the source image while stylizing them.
- − Loses the 'jealous' facial expression of the girlfriend on the right, making her look happy instead.
- − Changes the urban background to a rural landscape, failing to preserve the source setting.
OmniGen v2
- + Successfully preserves the urban setting and the bus from the original photo.
- + Maintains the relative facial expressions and character dynamics better than Model A.
- + Preserves the clothing and hair colors with high fidelity.
- − The art style is more generic modern anime than the specific Ghibli hand-painted aesthetic.
- − Lighting is flat and lacks the 'dreamy' quality requested in the prompt.
Verdict: FLUX.2 [klein] 9B followed the stylistic instructions much better, producing a beautiful watercolor illustration that truly feels like a Ghibli concept piece, though it failed to preserve the 'jealous' narrative and the urban setting. OmniGen v2 stayed much truer to the source image's content and narrative, but its art style is a basic vector-like anime style that lacks the requested texture and mood. FLUX.2 is the winner for its superior artistic execution of the transformation.
Golden Hour Stroll
Image Editing“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
AI Judge Analysis
FLUX.2 [klein] 9B
- + Excellent source preservation, maintaining identical facial features and clothing textures.
- + Realistic hair movement that fans out naturally.
- + Abundant, realistically detailed autumn leaves that fit the environment.
OmniGen v2
- + Effective hair motion that clearly flows in one direction.
- + Vibrant color palette.
- − Noticeable changes to the subject's face, losing the likeness of the source image.
- − The flying leaves are very sparse and look like simple digital stickers.
- − Background details like trees and flowers have been smoothed over and lost clarity.
Verdict: FLUX.2 [klein] 9B successfully executed the edit while maintaining high-fidelity source preservation, keeping the woman and dog identical to the original image. In contrast, OmniGen v2 significantly altered the woman's face and simplified the background, while the added leaves appeared artificial and lacked the requested dynamic energy.
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.2 [klein] 9B
- + Excellent typography with correct spelling and accents.
- + Clean vector emblem style with balanced composition.
- + Precisely matches requested color palette and texture.
- − The steam element is a bit thick and stylized compared to a realistic steam wisp.
OmniGen v2
- + Simple and elegant minimalist icon.
- + Good use of high contrast brown and cream tones.
- − Spelling error in the main text ('CAFFFLORIN' instead of 'Caffè Florian').
- − The banner and text arrangement is slightly cluttered at the bottom.
- − Lacks the subtle background texture requested in the prompt.
Verdict: FLUX.2 [klein] 9B followed all prompt instructions perfectly, including accurate spelling with the burial accent and a well-rendered 'Est. 1720' banner. OmniGen v2 failed on text accuracy and lacked the requested background texture, resulting in a less professional logo design.
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 [klein] 9B
- + Successfully follows the requested infographic flow with clear visual stages.
- + Modern vector aesthetic with a nice interpretation of the requested color palette.
- + Excellent detailed illustrations of the Saturn V and Lunar Module.
- − Numerous spelling errors in text labels (e.g., 'AOILLO', 'EARDHT', 'TRANSLURAL').
- − The logical flow of arrows is slightly confusing/disjointed.
OmniGen v2
- + Clean grid layout that feels like a professional corporate infographic.
- + Accurate adherence to the color palette constraints.
- − Failed to follow the prompt's mission (labeled as 'APOLO 17' instead of Apollo 11).
- − Icons are generic and do not match the specific mission stage descriptions (Saturn V, Lunar Module, etc.).
- − Significant text distortion and illegible labels.
Verdict: FLUX.2 [klein] 9B is the clear winner as it actually attempted to illustrate the specific steps of the Apollo 11 mission requested in the prompt, including the Saturn V and Lunar Module. While both models struggled with spelling, OmniGen v2 failed the core prompt instructions by generating a generic template for 'Apollo 17' with icons that did not represent the requested launch or landing vehicles.
Explore each model
Unified multimodal model for text-to-image generation, instruction-guided image editing, personalized generation, and virtual try-on