Black Forest Labs' aesthetically-tuned 12-billion parameter flow transformer optimized for high-quality images with incredible aesthetics, suitable for personal and commercial use
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.1 Krea [dev]
18.8
arena score
#48 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Imagen 4.0 Generate 001
17.1
arena score
#55 of 62 in Text-to-Image
Vote tally
Where the votes landed
FLUX.1 Krea [dev]
0%
win rate
Ties
0%
Imagen 4.0 Generate 001
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent realism with believable glass thickness and internal reflections.
- + Lighting accurately follows the request, coming softly from the left.
- + The glass cube has a realistic base and structural integrity.
- − The plant is slightly less 'partially visible through the glass' compared to the other model.
Imagen 4.0 Generate 001
- + Successfully positions the plant so it is clearly seen through the glass surfaces.
- + Clean, modern aesthetic with sharp edges.
- + Textured book cover adds a nice tactile detail.
- − The blue sphere appears to be levitating without physical support inside the cube.
- − The 'glass' cube behaves more like a collection of mirrors on certain faces, lacking transparency and depth.
Verdict: FLUX.1 Krea [dev] produces a much more realistic scene where the laws of physics are respected, showing a sphere resting inside a hollow cube with natural glass distortions. Imagen 4.0 Generate 001 provides a clean image but fails on physical logic, as the sphere is floating in the center of what appears to be a solid block rather than a hollow 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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Captures the motion blur from passing cars very effectively.
- + Realistic skin texture and age and lighting on the subject.
- + Authentic feeling of 'imperfect framing' requested in the prompt.
- − The subject is standing and holding the bike rather than actively repairing it.
- − The bicycle geometry is slightly warped near the pedal and chain area.
Imagen 4.0 Generate 001
- + Excellent adherence to the 'repairing' action with a tool in hand.
- + Superior facial detail and skin texture.
- + Strong composition with vibrant red/blue color contrast in rain reflections.
- − Lacks the requested motion blur for the passing cars.
- − The hands and fingers have slight anatomical merging issues around the tool.
Verdict: Imagen 4.0 provided a much better interpretation of the core action (repairing), whereas FLUX.1 Krea shown the man simply standing with the bike. While FLUX.1 Krea better captured the specific technical request for motion blur, Imagen 4.0's overall image quality, color depth, and engagement with the 'repair' prompt make it the superior choice.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Naturalistic hair and bead integration
- + Highly realistic skin texture and subtle battle damage
- + Excellent atmospheric Lighting and bokeh effects
- − The leather strap is slightly low on detail compared to the armor
Imagen 4.0 Generate 001
- + Intricate engraving on the plate armor
- + Clear depiction of the requested beads in braids
- + Strong use of warm lighting contrast
- − The beads/tubes in the hair look repetitive and artificial
- − Skin texture appears somewhat waxen or digitally smoothed
- − The torch flame has a slightly 'clip-art' feel in its rendering
Verdict: FLUX.1 Krea [dev] produces a significantly more lifelike and cinematic portrait with realistic skin, hair, and lighting integration. While Imagen 4.0 captures the specific details of the armor and braids well, the overall composition feels more like a 3D game asset than a photograph, whereas FLUX.1 achieves a convincing 'battle-worn' realism.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent professional typography and layout structure
- + Comprehensive food photo grid that looks like a real menu
- + Clean, high-contrast black and orange color scheme
- − Text consists of garbled, illegible characters
- − Repeats the 'Appetizers' heading twice
- − Images in the grid are slightly repetitive with many pizzas
Imagen 4.0 Generate 001
- + Very clean minimalist aesthetic with modern accents
- + Text is surprisingly legible and uses a bold sans-serif font as requested
- + Strong adherence to the grid layout with distinct sections for Appetizers, Pizza, and Mains
- − The layout is more of a graphic poster than a functional multi-item menu
- − Some geometric accents feel a bit disconnected from the restaurant theme
Verdict: Imagen 4.0 significantly outperforms FLUX.1 Krea by producing legible text and a much more creative, modern grid interpretation. While FLUX.1 Krea creates a standard menu format, the text is completely gibberish, whereas Imagen 4.0 follows all prompt instructions including specific category headings and high-quality food photography.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Clean, professional typography for the main title and sub-text.
- + Strong color contrast and high-quality rendering of the fire and glowing coals.
- + Very photorealistic texture on the burger bun and lettuce.
- − Failed to place the price in a starburst, opting for a plain circular text element instead.
- − The burger isn't truly 'exploded'; it looks more like a standard stack with a floating top bun.
- − Text rendering at the bottom is nonsensical gibberish.
Imagen 4.0 Generate 001
- + Excellent 'exploded' composition with ingredients dynamically suspended at angles.
- + Followed the instruction for the fiery glowing starburst for the price accurately.
- + Successfully integrated the fiery/glowing effect into the typography as requested.
- − The cheese has a slightly plastic, less photorealistic appearance compared to Model A.
- − The placement of the title over the bottom of the burger slightly obscures the image subject.
Verdict: Imagen 4.0 Generate 001 is the clear winner for its superior composition and literal adherence to the 'exploded burger' and 'fiery starburst' instructions. While FLUX.1 Krea produced a high-quality food photograph, it failed to create the requested sense of motion, whereas Imagen 4.0 successfully balanced all the complex prompt elements into a cohesive advertisement.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent realistic chalk texture with dusty smudges on the board.
- + Clean and visually appealing cursive handwriting style.
- + Realistic lighting and shadows within a café setting.
- − Hallucinated additional menu items like 'Gruffle Mushroom' rather than following the three-item list exactly.
- − Several spelling errors in the added text like 'coman' and 'uvucrations'.
Imagen 4.0 Generate 001
- + Perfectly spelled the main requested menu items, including the difficult 'Brown Butter Chocolate Chip Cookies'.
- + Captured the exact date 'APRIL 30, 2026' correctly in the title.
- + Accurately included the gluten-free footer text mentioned at the bottom of the prompt's intent.
- − Incorrectly included prompt instructions as literal text on the board (e.g., 'Tittle', 'Elegant chalk hand').
- − The handwriting looks more like a digital font than natural chalk.
- − The board texture is too flat and lacks the realistic chalk dust seen in the other image.
Verdict: FLUX.1 Krea [dev] produces a far more realistic image with convincing chalk textures and lighting, but it struggles with text accuracy and hallucinations. Imagen 4.0 Generate 001 is much better at rendering the specific text requested, but it fails significantly by including the meta-instructions of the prompt as literal text on the board. FLUX.1 Krea [dev] is preferred for a realistic mock-up, despite the spelling errors, as it avoids the catastrophic prompt leakage seen in Imagen 4.0.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
FLUX.1 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent photorealism with cinematic lighting and deep shadows.
- + Highly creative and surreal technical details on the horse's 'armor' or 'saddle'.
- + Clean, high-resolution aesthetic with a strong sense of scale against the planet.
- − Failed the logic test of the prompt; the astronaut is riding the horse, not the other way around.
- − The horse's front legs have some anatomical distortion where they connect to the tech block.
Imagen 4.0 Generate 001
- + Beautiful, ethereal art style with vibrant colors and cosmic details.
- + Dynamic composition with flowing mane and atmospheric trails.
- + High level of detail in the astronaut's suit textures and the horse's coat.
- − Failed the specific logic instruction; the astronaut is riding the horse instead of the horse riding the astronaut.
- − One of the horse's back legs blends awkwardly into the space cloud/dust.
Verdict: Both models failed the specific instruction to have the horse on top (horse riding the astronaut), instead providing the standard astronaut-on-horse interpretation. FLUX.1 Krea [dev] is preferred for its superior photorealism and unique mechanical details, whereas Imagen 4.0 offers a more traditional digital illustration look. Because both missed the core logic reversal, the winner is decided by the higher visual quality of FLUX.1 Krea [dev].
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent fur texture on the capybara
- + Mood lighting matches a real New York night taxi ride
- + High level of photo-realism in the materials
- − Includes an extra person in the passenger seat not requested
- − Capybara's cap looks like a construction helmet rather than a driver cap
- − Steering wheel position is awkward relative to the driver
Imagen 4.0 Generate 001
- + Perfect adherence to characters with only the driver and one passenger
- + Superior composition with a clear view from the front of the vehicle
- + The capybara's cap is a more accurate representation of a chauffeur/taxi cap
- − Capybara's paws look slightly reptilian or claw-like
- − The lighting on the businesswoman is a bit flat compared to the driver
- − Slightly less 'gritty' realism than image A
Verdict: Imagen 4.0 followed the prompt details much more accurately, correctly depicting a single businesswoman in the back seat whereas FLUX.1 added an extra passenger. While FLUX.1 has slightly more sophisticated lighting, Imagen 4.0's superior composition and attention to the specific requested details make it the better choice.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Strong vintage gothic aesthetic with dark parchment tones
- + Detailed, sprawling tree silhouettes that add to the atmosphere
- + Intricate web and thorn border design
- − Major text errors including 'Pasty Halloween' and repeated words in the scroll
- − Confused date formatting ('6.30.2026' instead of '30.10.2026')
- − Layout feels a bit cluttered with multiple competing scrolls
Imagen 4.0 Generate 001
- + Perfect text rendering for all requested details including the specific date and location
- + Highly legible and well-ordered layout
- + Atmospheric blue moonlight provides good contrast for the jack-o-lantern
- − Parchment effect on the left side looks like a rolled pole rather than a background texture
- − Slightly more 'digital illustration' feel compared to the requested 'vintage' style
- − The thorn border is a bit repetitive in its pattern
Verdict: Imagen 4.0 significantly outperformed FLUX.1 Krea by correctly rendering every piece of requested text, including the specific date and location, whereas FLUX.1 Krea suffered from significant spelling errors like 'Pasty Halloween'. While FLUX.1 Krea captured a more authentic vintage parchment texture, Imagen 4.0 is the superior choice for a functional invitation due to its clarity and accuracy.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent typography adherence including the requested text and flag icon.
- + Perfect lighting and material rendering for a clean 3D miniature look.
- + Matches the requested isometric perspective and light blue background exactly.
- − The sushi piece is very simple, showing only one type.
Imagen 4.0 Generate 001
- + High level of realism in the sushi textures, specifically the roe and avocado.
- + Clean composition on a miniature base.
- − Failed to include any of the requested text or the flag icon.
- − Background is grey/neutral instead of the requested light blue.
- − Missing the cartoon 3D aesthetic, opting for hyper-realism instead.
Verdict: FLUX.1 Krea [dev] followed the complex instructions perfectly, delivering the exact typography, background color, and isometric 3D style requested. Imagen 4.0 Generate 001 produced a realistic plate of sushi but ignored almost all specific formatting requests, including the text and background color.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent photographic lighting and depth of field
- + Realistic fur textures and animal anatomy
- + Captures a coherent sense of movement and playful energy
- − Completely missed the baby bunny requirement
- − Included two kittens instead of the requested variety
- − Butterflies appear slightly pasted on compared to the animal subjects
Imagen 4.0 Generate 001
- + Successfully included all four requested animals (puppy, kitten, bunny, fox)
- + Excellent depiction of dew sparkles and god rays
- + Captures a very whimsical and vibrant wildflower meadow
- − Less photorealistic and more illustrative/CG in style
- − Animal anatomy is a bit stiff, particularly the kitten's pose
- − Lighting feels more artificial compared to the naturalistic shadows in Model A
Verdict: FLUX.1 Krea [dev] produces a significantly more realistic and high-quality image with superior lighting and texture, but it failed the prompt instruction by omitting the bunny and doubling the kittens. Imagen 4.0 followed the prompt perfectly, including every requested animal and environmental detail, though it opted for a more stylized, digital art aesthetic rather than the requested hyper-photorealistic style.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Excellent woodcut engraving style suitable for a vintage logo.
- + High-quality banner texture and shading.
- + Includes the requested 'Est. 1720' date clearly.
- − Spells the name incorrectly as 'Café Flanelín'.
- − Distorted, nonsensical text at the bottom of the emblem.
Imagen 4.0 Generate 001
- + Perfectly adheres to the requested text 'Caffè Florian'.
- + Clean, minimalist vector aesthetic consistent with modern logo design.
- + Accurate representation of a cloche dome with steam.
- − The 'vintage' feel is much weaker than Model A, appearing more like a modern app icon.
- − The texture requested in the prompt is very subtle to the point of being almost absent.
Verdict: Imagen 4.0 is the clear winner for its perfect text rendering, accurately spelling 'Caffè Florian' which FLUX.1 Krea failed to do. While FLUX.1 Krea captured the vintage, textured engraving style much better, the incorrect spelling and hallucinated text make it unusable as a logo compared to the clean, accurate output of Imagen 4.0.
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 Krea [dev]
Imagen 4.0 Generate 001
AI Judge Analysis
FLUX.1 Krea [dev]
- + Successfully rendered the 'Apollo 11' header text clearly.
- + Uses a nice navy and lunar surface contrast.
- − Failed to include the Saturn V rocket or lunar module icons requested.
- − Iconography is nonsensical, including a planet that looks like Saturn and a person for step 4.
- − Text labels are garbled and the step numbers are disconnected from the visual meaning.
Imagen 4.0 Generate 001
- + Excellent typography with perfectly legible technical labels.
- + Much better adherence to the 'Saturn V' and 'Earth/Moon orbit' iconography requests.
- + Coherent and logical flow for an infographic.
- − Failed to include steps 5 and 6 (Descent and Landing).
- − The 'Earth Orbit' label points to a duplicate Earth icon instead of a distinct sequence.
Verdict: Imagen 4.0 significantly outperformed FLUX.1 Krea in terms of design coherence, text rendering, and style adherence. While Imagen 4.0 missed the final two steps of the mission, its professional vector aesthetic and accurate Saturn V illustration make it a much more successful infographic than FLUX.1 Krea, which produced logically disconnected icons and garbled text.
Next steps
Explore each model
Google's latest Imagen 4.0 text-to-image generation model with significantly better text rendering and overall image quality