Head to head
Esc

Models · slot A

to navigate to pick

FLUX.2 [dev] Flash fal Seedream 4.0 ByteDance

Settled by community votes across 20 shared challenges, with an AI judge weighing in on each.

FLUX.2 [dev] Flash

27.0 arena score

#7 of 62 in Text-to-Image

Skill signature · Text-to-Image

Seedream 4.0

24.7 arena score

#15 of 62 in Text-to-Image

Vote tally

Where the votes landed

FLUX.2 [dev] Flash

50.0%

win rate

Ties

16.7%

Seedream 4.0

33.3%

win rate

50.0% 16.7% ties 33.3%
Shared challenges 20

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Perfect prompt adherence for all spatial relationships.
  • + Excellent handling of glass textures and internal reflections.
  • + High level of photographic realism with natural lighting and depth of field.
  • The glass cube is slightly dusty/noisy, though this adds to realism.

Seedream 4.0

  • + Strong adherence to the core prompt elements.
  • + Vibrant colors on the red book and blue sphere.
  • + Good capture of soft window light.
  • Perspective issues on the top glass pane where it meets the book.
  • The base of the cube looks like a mirror rather than a simple glass pane on a table.
  • Lower resolution feel compared to the other model.

Verdict: FLUX.2 [dev] Flash produced a significantly more realistic image with superior handling of physics, such as how the glass edges interact with the light and the book. Seedream 4.0 followed the prompt well but struggled with the structural integrity of the cube's top and used a mirrored base that wasn't requested.

Man and Car in California

Editing
Edit instruction

“Make a photo of the man driving the car down the California coastline”

Source
FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent preservation of the man's facial features and clothing.
  • + Realistic interior detail and lighting integration.
  • + High cinematic quality with natural depth of field.
  • The car interior changed from the source's white/light theme to a wood-grain vintage style.
  • The car's exterior perspective was changed significantly.

Seedream 4.0

  • + Successfully preserved the specific Rolls Royce exterior model from the source image.
  • + Accurately placed the subject in a California coastline setting with motion blur.
  • Loss of facial likeness, making the man look different from the source.
  • Rough blending around the subject's hair and the car's windshield frame.

Verdict: FLUX.2 [dev] Flash delivered a more realistic and high-quality image, maintaining the subject's likeness perfectly, though it altered the car's interior design. Seedream 4.0 preserved the original car model better but failed to maintain the man's facial features and had poorer technical blending around the edges of the subject.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent realism with highly detailed skin textures and lifelike hands.
  • + Superior lighting and reflections on the wet pavement that match the rainy atmosphere.
  • + Accurate motion blur on the background cars while keeping the subject sharp.
  • The bicycle mechanics are slightly nonsensical, with some frame bars and cables merging awkwardly.
  • The 50mm lens effect feels slightly flatter than requested relative to the background.

Seedream 4.0

  • + Good adherence to the 'shallow depth of field' and 'motion blur' requests.
  • + The composition feels very much like a 'candid street photo' with imperfect framing.
  • The face suffers from significant deformation and lack of clear detail.
  • The bicycle spokes and frame have severe rendering artifacts and structural breaks.
  • General image quality is lower with a slightly digital/blurry film grain texture.

Verdict: FLUX.2 [dev] Flash is the clear winner due to its exceptional rendering of human textures, wet surfaces, and atmospheric lighting. While Seedream 4.0 followed the composition cues of the prompt well, it failed significantly on technical execution, resulting in a distorted face and broken bicycle geometry.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Exceptional sharpness and detail in the engraved plate armor and leather textures.
  • + Very clean application of the beads in the braids.
  • + Excellent symmetry and balanced composition with realistic flickering torchlight bokeh.
  • The scars look a bit like fresh red paint or stage blood rather than healed battle marks.

Seedream 4.0

  • + Stronger sense of 'warm torchlight' with a more dramatic high-contrast lighting scheme.
  • + Scars and dirt are blended more naturally into the skin texture.
  • + Dynamic composition with a slightly angled head pose.
  • Overall lower image resolution and softer details on the armor engraving.
  • The hair braids are significantly less detailed and have fewer beads than requested.
  • Some awkward lighting artifacts on the right side of the face.

Verdict: FLUX.2 [dev] Flash is the clear winner due to its superior technical execution, particularly in the intricate engraving of the armor and the crispness of the leather straps and hair beads. While Seedream 4.0 captures a more cinematic lighting mood, it struggles with the fine textures and resolution required by the prompt.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent structure that follows a real menu layout including pricing and headers.
  • + Accurate adherence to the grid-style photo request with vibrant color-coded accents.
  • + Uses bold sans-serif fonts appropriately for a professional casual dining look.
  • Text rendering contains severe misspellings and gibberish typical of some AI models.
  • The section logic is slightly flawed, as the 'Pizza' header is placed next to text rather than above photos.

Seedream 4.0

  • + Very clean and legible text for the main section headers.
  • + High-quality, appetizing food photography with good clarity.
  • + Captures the 'minimalist' aesthetic well with large negative space.
  • Fails to include actual menu items, descriptions, or prices.
  • Layout is less like an actual menu and more like a collage or mood board.
  • The grid is uneven and lacks the professional structure implied by the prompt.

Verdict: FLUX.2 [dev] Flash successfully created a comprehensive menu layout with headers, pricing, and organized sections, although the text is mostly unreadable. Seedream 4.0 produced higher quality individual food images and very clean heading text, but completely failed to generate a functional menu design, leaning instead into a simple photo collage. FLUX.2 [dev] Flash is the winner for following the specific structural requirements of a professional menu.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent typography rendering with consistent fiery glow effects.
  • + Highly detailed and photorealistic food textures, especially the patty and bun.
  • + Perfect adherence to the specified price point and starburst prompt.
  • The 'exploded' effect is slightly static compared to the swirl motions in the competitor.
  • The lighting on the bottom bun feels a bit disconnected from the dark ground.

Seedream 4.0

  • + Dynamic sense of motion with swirling trails around the food components.
  • + Strong 'magic' atmosphere with more aggressive fire and ember effects.
  • + Composition creates a more three-dimensional feel.
  • Failed to render the correct price, showing €5.99 instead of €6.99.
  • The burger is split into two mini-burgers rather than one exploded burger.
  • Lower texture realism on the vegetable components compared to Model A.

Verdict: FLUX.2 [dev] Flash followed the prompt instructions precisely, accurately rendering the text and price while maintaining high photorealistic detail. Seedream 4.0 offered a more creative and dynamic composition with swirl effects, but it failed on the specific price requirement and altered the burger structure.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent text rendering with perfect spelling across all items.
  • + Clean and legible composition that feels very realistic for a cafe setting.
  • + High fidelity chalk texture on the individual letters.
  • The 'Today's Specials' heading is in a print-style block font rather than the requested elegant cursive.
  • The background of the board looks a bit too clean and digitally filled in some areas.

Seedream 4.0

  • + Successfully used elegant cursive for the 'Today's Specials' title as requested.
  • + Very realistic chalkboard texture with dust, smudges, and authentic chalk residue.
  • + Excellent adherence to the 'handwritten' feel with varying line weights.
  • Slightly less crisp text legibility compared to Model A.
  • The chalk dust at the bottom looks a bit clumped and digitally generated upon close inspection.

Verdict: Both models performed exceptionally well on a difficult text-heavy prompt. FLUX.2 [dev] Flash delivered cleaner text and perfect spelling, but Seedream 4.0 followed the specific stylistic instruction for cursive titles more accurately and provided a more convincing, messy chalkboard atmosphere. Seedream 4.0 is the overall winner for better adherence to the specific 'cursive' and 'textured' qualitative descriptors in the prompt.

Pose & Character Mashup

Editing
Edit instruction

“Use Image 1 as the exact pose reference and Image 2 as the character reference. Recreate the person/character from Image 2 in the exact dynamic pose and body position from Image 1. Keep the exact face, hair, clothing style/details, and expression from Image 2. Match the lighting and environment of Image 1. The final image must show the character from Image 2 performing the precise action/pose from Image 1 with perfect anatomy and natural integration.”

Source
FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent character replication including the sunglasses, scarf, and facial features from Image 2.
  • + Maintains high resolution and realistic textures relative to the source.
  • + Successfully replicates the lighting and background of Image 1.
  • Failed to match the body position, creating a standard crouching pose instead of the dynamic bent pose from Image 1.
  • Major anatomical error with an extra head/hair bundle appearing behind the main character.

Seedream 4.0

  • + Closer attempt at the specific leaning/dynamic body angle of Image 1.
  • + Incorporates the scarf and clothing style requested from Image 2.
  • Completely failed to replicate the character's face, sunglasses, and short hair from Image 2.
  • Anatomical issues with the feet and a distorted face that lacks clarity.
  • Failed to capture the specific 'character reference' entirely.

Verdict: Both models struggled significantly with the complex instruction. FLUX.2 [dev] Flash successfully captured the identity of the person in Image 2 (face, hair, glasses) but failed the pose and produced a scary artifact of a second head. Seedream 4.0 attempted the pose slightly better but failed to use the character from Image 2 at all, instead mixing traits from both images into a low-quality figure.

The Reversed Rodeo

Text-to-Image

“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”

FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent high-detail rendering of the spacesuit and horse texture.
  • + Great cinematic composition with multiple celestial bodies and a planetary horizon.
  • + Consistent lighting and clear, sharp resolution.
  • Failed to follow the specific spatial instruction 'horse on top, not vice versa'.

Seedream 4.0

  • + Dynamic, energetic posing of the horse.
  • + Vibrant colors in the nebula and helmet reflection.
  • + Good adherence to the 'surreal' aspect of the prompt.
  • Failed to follow the specific spatial instruction 'horse on top, not vice versa'.
  • Minor artifacts where the astronaut's hand meets the reins.
  • Less depth in the background compared to Image A.

Verdict: Both models completely ignored the negative constraint/instruction to place the 'horse on top' of the astronaut, resulting in standard images of astronauts riding horses. While both failed the main logic test, FLUX.2 [dev] Flash is the superior image due to its much higher level of detail, better orbital composition, and more realistic textures.

Outfit Transfer Challenge

Editing
Edit instruction

“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”

Source
FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Successfully captured the peacoat and plaid scarf elements.
  • + Added elaborate jewelry details as requested in the prompt.
  • Extreme facial distortion and anatomical errors on the subject's face.
  • Ghosting effect with a second head appearing behind the main subject.
  • Failed to preserve the original person's face accurately.

Seedream 4.0

  • + Successfully transferred the pea coat, scarf, jeans, and sunglasses from Image 2.
  • + Maintained the overall composition and background of Image 1.
  • + Reasonably preserved the vitiligo pattern from the original subject.
  • The face is a blend of both models rather than the 'exact face' from Image 1.
  • Noticeable artifacts around the ankles and shoes.
  • The lighting on the subject is slightly inconsistent with the beach background.

Verdict: FLUX.2 [dev] Flash fails the task due to severe facial distortion and anatomical glitches, including a secondary head. Seedream 4.0 successfully performs the clothing transfer while maintaining the pose and background, though it struggles with perfectly preserving the subject's original facial features 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.”

FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent photorealism with sharp textures on the capybara's fur and the businesswoman's coat.
  • + Higher quality lighting and bokeh in the background, creating a convincing Manhattan night atmosphere.
  • + Dynamic and centered composition that perfectly captures the interaction between the driver and passenger.
  • The capybara's paws look slightly more like claws, which is anatomically accurate but visually jarring.
  • The taxi 'TX' sign on the roof is slightly oddly placed relative to the windshield.

Seedream 4.0

  • + Perfectly captures the 'bored' expression requested for the passenger.
  • + Good side-profile perspective of the capybara driving.
  • + Adheres well to all primary prompt elements like the yellow cap and jacket.
  • Noticeable anatomical issues with the capybara's right paw, which appears mangled or poorly fused with the steering wheel.
  • Overall image clarity is softer and less detailed than its competitor.
  • Text rendering on the taxi door and cap is nonsensical.

Verdict: FLUX.2 [dev] Flash is the winner due to its superior technical quality and photorealistic rendering. While Seedream 4.0 followed the prompt well, it suffered from significant anatomical artifacts on the capybara's hands and lacked the sharp detail present in the FLUX.2 image.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent text legibility and spelling throughout.
  • + Cohesive and symmetrical composition that feels like a professional poster.
  • + Includes all requested elements like thorns, webs, and scroll banners accurately.
  • Includes a line of gibberish text 'Hurk: 0369' above the event time.
  • The background trees are somewhat repetitive in shape.

Seedream 4.0

  • + Atmospheric cinematic lighting on the jack-o-lantern.
  • + The torn parchment aesthetic adds a tactile, vintage feel to the borders.
  • + Dynamic placement of bats and twisted trees.
  • Text on the banner is distorted and difficult to read.
  • Composition feels slightly cluttered with the large folded parchment in the bottom right corner.
  • Typography at the bottom is less polished than Model A.

Verdict: FLUX.2 [dev] Flash produced a much more functional invitation with clear, legible text and a structured layout, despite adding one line of hallucinated text. Seedream 4.0 has better atmospheric lighting and texture, but fails on the legibility of the small scroll banner. FLUX.2 is the winner for its superior graphic design quality and clear information delivery.

Bald man challenge

Image Editing
Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After
FLUX.2 [dev] Flash
Before After
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent preservation of the original face, clothing, and background.
  • + Highly realistic hair texture with individual strands visible against the sky.
  • + Natural hairline integration that respects the original forehead structure.
  • The 'afro' style chosen is a significant departure from the original hair texture suggested by his beard.

Seedream 4.0

  • + Successful addition of a full head of hair that matches the requested density.
  • + Good preservation of the source image's identity and lighting.
  • The hairline looks slightly artificial and 'stamped on' near the forehead.
  • The hair texture appears a bit fuzzy or soft compared to the sharpness of the beard.

Verdict: Both models did an excellent job of preserving the source image while adding hair. FLUX.2 [dev] Flash is the winner because the hair texture is significantly more realistic and detailed, with fine individual hairs blending naturally into the background, whereas Seedream 4.0's hair looks slightly more like a digital painting or wig.

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.2 [dev] Flash
Seedream 4.0
25% wins 25% ties 50% wins

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Perfect text rendering of 'JAPAN' and 'SUSHI'
  • + Excellent isometric perspective and clean diorama base
  • + High-quality soft textures and realistic lighting
  • The sushi roll construction is slightly illogical with fish draped over a sliced roll
  • Minimal variety in the sushi types presented

Seedream 4.0

  • + Good variety of sushi types including nigiri, maki, and gunkan
  • + Accurate text and flag icon placement
  • + Appealing miniature aesthetic with a more traditional wooden sushi board
  • Slightly lower clarity in the background and shadows compared to Model A
  • Text is a bit tighter and less balanced in the top margin

Verdict: Both models followed the prompt exceptionally well, producing clean 3D isometric designs with accurate text. FLUX.2 [dev] Flash is the winner because its rendering is noticeably sharper, the lighting is more professional, and it perfectly captured the 'ultra-clean' aesthetic requested in the prompt, despite a slightly odd sushi arrangement.

Over-the-top cartoon caricature

Editing
Edit instruction

“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”

Source
FLUX.2 [dev] Flash
Seedream 4.0

AI judge analysis unavailable for this challenge.

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.2 [dev] Flash
Seedream 4.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent fur texture and detail across all animals.
  • + Includes all requested animals clearly with a high degree of photorealism.
  • + Strong composition with a balanced arrangement of animals and butterflies.
  • Generated an extra bunny, which wasn't specifically requested but fits the theme.

Seedream 4.0

  • + Captures the 'tumbling' and 'playful' motion dynamic very well.
  • + Effective use of backlighting and sparkling dew drops.
  • + Good adherence to the specific animal types requested.
  • Anatomical issues with the fox's legs and paws.
  • The kitten's tail and hind area are poorly defined against the background.
  • Slightly less 'hyper-photorealistic' than Model A, appearing more like a digital painting.

Verdict: FLUX.2 [dev] Flash delivers a much cleaner and more detailed image with superior anatomical accuracy and fur rendering. Seedream 4.0 captures the movement and light of the prompt well but suffers from noticeable AI artifacts in the limbs and blending of the animals. FLUX.2 is the clear winner for its polish and realism.

Studio Ghibli Anime Style

Editing
Edit instruction

“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”

Source
FLUX.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent preservation of the original pose, clothing, and facial expressions.
  • + Incorporates charming Ghibli-esque details like the Totoro-like figure in the background.
  • + Matches the specific warm, nostalgic lighting and color palette requested.
  • The line work is a bit busier than typical Ghibli character designs.
  • The eyes are slightly more 'generic anime' than the distinctive Ghibli style.

Seedream 4.0

  • + Authentic watercolor texture that feels like traditional hand-painted background art.
  • + Simplifies facial features in a way that is highly characteristic of Ghibli character design.
  • Loses the specific pained expression of the girl on the right, making her look bored instead.
  • The background is very abstract and loses the structural context of the original scene.

Verdict: FLUX.2 [dev] Flash is the clear winner as it perfectly translates the world-famous 'Distracted Boyfriend' meme into an illustration while maintaining all key narrative elements, including the specific facial expressions. Seedream 4.0 provides a beautiful watercolor texture, but it fails to capture the emotional intent of the scene and simplifies the background too much.

Golden Hour Stroll

Image Editing
Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

Before After
FLUX.2 [dev] Flash
Before After
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Stronger adherence to the wind-blown hair request with more volume and spread
  • + Maintains high resolution and sharpness from the original source
  • + Adds a high quantity of leaves that match the environment's palette
  • The hair flow is a bit symmetrical and looks slightly unnatural
  • Numerous small leaf particles appear scattered and messy in some areas

Seedream 4.0

  • + Realistic interpretation of hair movement that looks natural and windswept
  • + Effective use of motion blur on the falling leaves to convey speed
  • + Excellent preservation of the subject's face and the dog while adding atmosphere
  • Slightly lower overall clarity compared to Model A
  • Adds a lens flare/lighting effect in the top right that was not requested

Verdict: FLUX.2 [dev] Flash and Seedream 4.0 both successfully edited the image to add motion. Seedream 4.0 is the winner as its motion effects feel more integrated and realistic, particularly the natural hair flow and the motion blur on the leaves, whereas FLUX.2 [dev] Flash produced more static-looking hair and a high volume of sharp, cluttered leaf particles.

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.2 [dev] Flash
Seedream 4.0

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Excellent typography with a balanced banner design
  • + Professional vector emblem style with cohesive texture
  • + Accurate rendering of the accent mark in 'Caffè'
  • The 'Est. 1720' text is placed below the banner rather than inside it

Seedream 4.0

  • + Successfully placed 'Est. 1720' on the banner as requested
  • + Clean use of warm brown tones and paper texture
  • The main brand name is floating awkwardly above the icon rather than being integrated
  • The cloche is slightly lopsided and lacks the refinement of a vector logo

Verdict: FLUX.2 [dev] Flash produces a much more professional and coherent logo design that feels like a real brand identity, despite placing the date below the banner. Seedream 4.0 follows the banner instruction more literally but lacks the compositional balance and refined typography found in FLUX.2.

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.2 [dev] Flash
Seedream 4.0
100% wins 0% ties 0% wins

AI Judge Analysis

FLUX.2 [dev] Flash

  • + Captures the flat-vector aesthetic with sophisticated textures and a modern NASA-inspired layout.
  • + Excellent adherence to the color palette requests and detailed illustration of the Lunar Module.
  • + Very high resolution for the icons and text, making it look like a professional poster design.
  • Layout is somewhat cluttered with redundant labels for 'Descent' and 'Landing'.
  • Includes gibberish text in the 'Saturn V Icon' label and supporting text blocks.

Seedream 4.0

  • + Logical, linear timeline layout that is very easy to follow as an infographic.
  • + Clean, bold typography that is mostly legible and correctly indexed 1-6.
  • + Strict adherence to the 6-step structure requested in the prompt.
  • The art style is a bit too simplistic and cartoonish compared to the requested 'modern' vector look.
  • Text clipping occurs on 'Translunar' (cut off as 'Transl lunar') and 'Tranquility'.
  • Low-quality rendering on some icons, such as the Moon and the landing craft.

Verdict: FLUX.2 [dev] Flash produced a much more visually appealing and professional poster with high-quality vector illustrations and a sophisticated NASA-inspired palette. While Seedream 4.0 followed the 6-step linear structure more logically, its art style is overly simplistic and it suffered from text-clipping issues that diminished its quality as an infographic.

Next steps

Explore each model