Black Forest Labs' distilled 9 billion parameter image generation model with sub-second inference and multi-reference support
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.2 [klein] 9B
#13 of 32 in Image Editing
Imagen 4.0 Generate 001
#55 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [klein] 9B
0%
win rate
Ties
0%
Imagen 4.0 Generate 001
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 adherence to the 'partially visible through the glass' instruction for the plant.
- + Realistic glass refraction and surface imperfections on the wooden table.
- + The cube shape is well-defined and accurately holds the book on its flat top.
- − The blue sphere appears to be sitting inside a glass vase/vessel rather than a solid or hollow cube.
- − Distortions on the left side of the glass look slightly messy.
Imagen 4.0 Generate 001
- + Perfect geometric cube shape with very clean edges.
- + Excellent lighting consistency across the book, sphere, and cube surfaces.
- + The blue sphere is perfectly centered and rendered with a nice matte finish.
- − The plant behind the cube is completely obscured by the glass's reflections/refractions, failing that part of the prompt.
- − The glass appears more like a mirror or lead crystal rather than simple clear glass.
Verdict: FLUX.2 [klein] 9B followed the complex spatial instruction regarding the plant being visible through the glass much better than Imagen 4.0, which rendered the glass as nearly opaque or highly reflective. While Imagen 4.0 produced a cleaner geometric cube, FLUX.2 [klein] 9B created a more believable and photographic scene that adhered to all prompt details.
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 realism in skin texture and fabric materiality.
- + Highly detailed and accurate bicycle and tool kit components.
- + Superior lighting and depth of field that feels like a genuine photograph.
- − The motion blur on the passing car is quite subtle, leaning more towards a static shot.
- − Strictly follows a clean composition despite the 'imperfect framing' prompt.
Imagen 4.0 Generate 001
- + Stronger sense of 'imperfect framing' with the subject pushed more to the center-left and the bike cut off.
- + Good usage of colors and reflections on the wet pavement.
- + The pose captures a more active sense of 'repairing' with a tool in hand.
- − Anatomical issues with the hands, including a missing finger on the right hand.
- − Non-sensical mechanical details where the tool meets the bicycle chain/derailleur area.
- − Image quality feels slightly more processed and less like a raw 50mm photograph compared to its counterpart.
Verdict: FLUX.2 [klein] 9B produces a significantly more realistic and technically sound image, with excellent skin textures and a believable mechanical bicycle. While Imagen 4.0 Generate 001 captures the 'imperfect framing' and 'candid' feel slightly better, it suffers from noticeable AI artifacts in the hands and bicycle components that break the immersion.
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 depiction of scars and dirt on the skin.
- + Highly detailed and realistic leather strap texture with visible wear.
- + Accurate braided hair with distinct bead placement.
- − The torch in the background is slightly sharp, reducing the 'shallow depth of field' effect compared to the subject.
Imagen 4.0 Generate 001
- + Strong cinematic lighting with warm torchlight highlights across the face.
- + Intricate engraving details on the armor plates.
- + Good use of bokeh sparks and atmospheric depth.
- − The hair beads look more like mechanical clips than decorative beads.
- − Skin texture and scars appear slightly more stylized and less lifelike than Model A.
Verdict: FLUX.2 [klein] 9B provides a more realistic and gritty interpretation of a battle-worn paladin, with superior skin textures and authentic-looking leather. Imagen 4.0 Generate 001 offers a more cinematic, 3/4 view composition with impressive armor engravings, but the hair details look slightly artificial. FLUX.2 is preferred for its better adherence to the physical descriptions of scars, dirt, and lifelike textures.
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 adherence to the 'sections' instruction with clear headers for Appetizers, Pizza, and Mains
- + Professional font choice and high-quality food photography
- + Cleaner overall layout that resembles a real-world vertical menu
- − Repeats section labels ('Appetizers' twice, 'Mains' twice)
- − Text becomes messy and glitched in the lower right section
Imagen 4.0 Generate 001
- + Stronger adherence to the 'grid' layout requirement
- + More vibrant color accents using geometric shapes
- + Highly consistent image quality across all food photos
- − The grid arrangement makes it harder to follow as a functional menu
- − Legibility of text in the small boxes is slightly lower than the main headers of the other model
Verdict: Both models followed the prompt well, but FLUX.2 [klein] 9B produced a more realistic restaurant menu structure despite some text repetition and glitches. Imagen 4.0 provided a superior grid layout with better color accents, but it felt more like a social media graphic than a functional menu. FLUX is preferred for its professional compositional logic.
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 typography with a cohesive 'fiery' treatment across all text elements.
- + High-quality, appetizing food textures with realistic lighting.
- + Strong sense of dynamic motion with sauce droplets and sparks radiating from the center.
- − The burger is less 'exploded' than Model B, appearing more as a slightly open stack.
Imagen 4.0 Generate 001
- + Excellent 'exploded' composition with clear space between all ingredients.
- + Accurate adherence to the request for a starburst price tag and individual suspended components.
- + High clarity and sharp focus on the floating vegetables and sauce.
- − The primary text lacks the requested 'fiery, glowing effect' compared to Model A.
- − The texture of the cheese appears slightly plastic or synthetic under harsh lighting.
Verdict: FLUX.2 [klein] 9B produces a more professional-looking advertisement with superior text integration and lighting, though the burger is less deconstructed. Imagen 4.0 Generate 001 follows the 'exploded' mechanical prompt more literally, making every ingredient distinct, but falls short on the artistic rendering of the fiery text. FLUX.2 [klein] 9B is the preferred choice for a commercial ad aesthetic.
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 no spelling errors in main items.
- + High realism with authentic chalk textures and eraser smudges.
- + Successfully captured the elegant cursive style requested for the title.
- − Small typo in the footer text: 'fress' instead of 'fresh'.
Imagen 4.0 Generate 001
- + Successfully rendered the requested menu items and prices.
- + Good framing with the wooden border.
- − Included meta-instructions from the prompt as literal text on the board.
- − Numerous spelling errors and gibberish text in the supplemental sections.
- − The handwriting looks more like a digital font than natural chalk.
Verdict: FLUX.2 [klein] 9B is the clear winner as it produced a highly realistic, legible, and aesthetically pleasing chalkboard menu. Imagen 4.0 Generate 001 failed significantly by printing the prompt's descriptive instructions (like 'Tittle Menu' and 'Footer') directly onto the image alongside nonsensical text.
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 depiction of multiple celestial bodies in the background
- + Very realistic horse anatomy and muscle definition
- + Strong cinematic lighting consistent with the environment
- − Includes nonsensical AI-generated text at the bottom
- − The prompt specifically asked for 'horse on top', which could be interpreted as a spatial reversal, though it followed the literal riding instruction.
Imagen 4.0 Generate 001
- + Creative integration of the cosmos into the horse's coat
- + Better flow and movement in the horse's mane and tail
- + Cleaner image without distracting artifacts or gibberish text
- − The transition between the horse's feet and the atmosphere looks slightly blurred and less defined
Verdict: Both models failed the negative constraint/inverse logic of 'horse on top, not vice versa', likely prioritizing the 'horse riding' action phrase. Imagen 4.0 is the superior choice because it produced a much cleaner, more artistic image without the distracting gibberish text found in FLUX.2.
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 photorealistic texture on the capybara's fur and the passenger's face.
- + Captures the bored, non-chalant expression of the businesswoman perfectly.
- + Realistic lighting from the phone screen illuminating the passenger's hands and face.
- − Placed the passenger in the front seat instead of the back seat as requested.
- − The capybara's paws look slightly distorted and more like human-gloved hands.
Imagen 4.0 Generate 001
- + Followed the spatial instructions correctly by placing the passenger in the back seat.
- + Accurate rendering of the capybara's anatomy, specifically the large claws/paws on the steering wheel.
- + Clear legible text on the driver's cap and the taxi sign.
- − The passenger's expression looks slightly too focused/intense rather than 'completely normal' or 'bored'.
- − The overall image has a slightly more processed, CGI-like finish compared to the photorealism of Model A.
Verdict: While FLUX.2 [klein] 9B achieved a more impressive level of photorealism and captured the requested 'bored' expression better, it failed the core compositional instruction of placing the passenger in the back seat. Imagen 4.0 Generate 001 correctly placed the businesswoman in the back and maintained better anatomical accuracy for the capybara's feet/paws, making it the more successful image overall for this specific prompt.
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
- + Perfect text rendering for all requested strings, including the date and locations.
- + Exquisite gothic typography that blends naturally with the vintage aesthetic.
- + Central jack-o-lantern has superior cinematic lighting and texture.
- − The parchment edge is a bit clean, though it does feature a subtle aged texture.
Imagen 4.0 Generate 001
- + Good use of negative space for the bottom event details.
- + Innovative inclusion of a physical-looking scroll element on the left side.
- − Text rendering has minor stylistic inconsistencies in the font weight.
- − The 'parchment' look is less convincing, feeling more like a digital graphic than a vintage poster.
- − The thorn border is somewhat repetitive and less integrated than in the competitor's image.
Verdict: FLUX.2 [klein] 9B followed the prompt perfectly, producing a highly polished and atmospheric invitation with flawless typography. While Imagen 4.0 provided a creative layout with the side-scroll, its overall visual style feels more like a modern vector illustration rather than the requested 'vintage' and 'dark parchment' aesthetic.
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 adherence to typographic instructions including central text and flag icon.
- + Correctly applied the requested light blue background and isometric diorama base.
- + Achieved a clean, consistent 3D cartoon art style.
- − The flag icon is incorrect, showing three horizontal stripes instead of a red circle.
- − The sushi models are slightly simplistic compared to the complex PBR request.
Imagen 4.0 Generate 001
- + High visual quality with realistic textures on the rice and fish.
- + Good lighting and shadows that create a miniature feel.
- + Followed the diorama base and circular plate instructions.
- − Completely ignored all typographic instructions including the text 'JAPAN' and 'SUSHI'.
- − Failed to provide a solid light blue background.
- − Is missing the requested flag icon.
Verdict: FLUX.2 [klein] 9B is the clear winner as it followed almost every specific instruction in the prompt, including the complex text and layout requirements, despite the flag being inaccurate. Imagen 4.0 provided a more realistic texture but failed to include the requested text, background color, and icons, making it insufficient for the specific brief.
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 rendering with realistic fur texture
- + Sophisticated lighting with natural god rays and bokeh
- + Cohesive composition where the animals feel grounded in the same space
- − Missed the baby bunny from the prompt entirely
- − The kitten has an oddly elongated neck/torso
Imagen 4.0 Generate 001
- + Successful inclusion of all four animals requested: puppy, kitten, fox, and bunny
- + Captures the 'tumbling together' action more literally than Image A
- + Includes prominent dew sparkles on the grass as requested
- − Visual style leans towards illustration/digital art rather than hyper-photorealism
- − The kitten is missing one of its front legs in its pose
- − Animal proportions and lighting feel a bit composited rather than natural
Verdict: While FLUX.2 [klein] 9B produces a much more realistic and aesthetically pleasing image with superior lighting, it failed to include the bunny. Imagen 4.0 Generate 001 provides all requested subjects and captures the 'tumbling' interaction better, but it lacks the photographic realism requested and has some anatomical flaws in the kitten.
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 Italian diacritics
- + Effective use of the banner and emblem circular frame
- + Superior texture on the background that fits the vintage theme
- − The steam design is a bit thick and looks slightly cartoonish
Imagen 4.0 Generate 001
- + Clean minimalist vector style with balanced proportions
- + Accurate representation of a cloche with subtle steam lines
- + Good adherence to the warm brown and cream color palette
- − The background lacks the requested subtle texture
- − The 'Est. 1720' banner is small and detached from the main emblem structure
Verdict: FLUX.2 [klein] 9B followed the prompt more closely by integrating all elements into a cohesive vector emblem with a textured background. While Imagen 4.0 Generate 001 produced a very clean modern-minimalist logo, FLUX.2 captured the 'vintage' aspect better through its typography and layout.
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 includes icons for both descent and landing phases.
- + Good use of custom iconography for the crew and lunar module.
- − Numerous spelling errors in text (e.g., 'AOILLO', 'EARDHT', 'TRANSLURAL').
- − The infographic flow is disorganized and lacks a clear logical path.
- − Includes inconsistent design choices like the green on Earth which clutters the navy/red palette.
Imagen 4.0 Generate 001
- + Perfect text rendering with zero spelling errors.
- + Clean, logical infographic layout with a clear trajectory path.
- + High adherence to the requested NASA-inspired color palette and flat-vector style.
- − Misses the final two specific icons for 'Descent' and 'Landing'.
- − The Saturn V rocket is slightly cropped at the bottom.
Verdict: Imagen 4.0 Generate 001 is the superior choice because it produces a professional, clean vector graphic with perfect typography and a logical layout. While FLUX.2 [klein] 9B includes more of the requested steps, its execution is marred by significant spelling mistakes and a confusing visual flow.
Explore each model
Google's latest Imagen 4.0 text-to-image generation model with significantly better text rendering and overall image quality