Google's latest Imagen 4.0 text-to-image generation model with significantly better text rendering and overall image quality
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
Imagen 4.0 Generate 001
17.1
arena score
#55 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Seedream 4.5
26.1
arena score
#9 of 62 in Text-to-Image
Vote tally
Where the votes landed
Imagen 4.0 Generate 001
20.0%
win rate
Ties
0.0%
Seedream 4.5
80.0%
win rate
20.0%
0.0% ties
80.0%
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent photorealism in the book texture and table wood grain
- + Sharp focus and high resolution
- + Consistent lighting across all surfaces
- − The cube appears to be a solid block of glass/acrylic rather than a hollow cube, causing the sphere to look embedded rather than 'inside'
- − The plant is entirely behind the object rather than visible through it as requested
Seedream 4.5
- + Successfully renders a hollow glass cube with visible interior space
- + Correctly shows the plant partially visible through the glass panels
- + Includes a realistic shadow and reflection for the blue sphere
- − The green plant is slightly blurry compared to the foreground
- − The lighting on the book is a bit flat compared to the table surface
Verdict: Seedream 4.5 captures the spatial relationships of the prompt much better, specifically the 'hollow' nature of the cube and the visibility of the plant through the glass. While Imagen 4.0 has higher textural fidelity, it failed to interpret the cube as a container, making the sphere look like a marble suspended in solid glass.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent skin texture and facial details
- + Beautiful rain drop details on clothing and bicycle
- + Strong bokeh effect that creates a cinematic look
- − Missed the request for motion blur on passing cars
- − The bicycle mechanics are slightly nonsensical around the chain area
- − Less realistic 'candid' feel compared to the other model
Seedream 4.5
- + Perfectly captures the motion blur of the passing car as requested
- + Composition feels more 'candid' and aligned with street photography
- + Logical reflections in the puddle and wet pavement
- − Facial features are less detailed and slightly muddy
- − Raindrops on the hood are oversized and look somewhat digital
- − Low resolution or slight softness on the subjects face compared to Model A
Verdict: Seedream 4.5 followed the complex technical prompt more accurately, specifically by including the requested motion blur of passing cars which Imagen 4.0 failed to do. While Imagen 4.0 produced a much more detailed and aesthetically pleasing portrait with superior skin texture, Seedream 4.5 captured the requested 'candid' street photography atmosphere and imperfect framing more effectively.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent clarity on the ornate engraving of the plate armor.
- + Clear, visible beads within the braided hair.
- + Balanced lighting that clearly defines the character's features.
- − The scars look a bit more like surface drawing than deep tissue marks.
- − Leather straps and buckles appear slightly flat and less textured compared to the armor.
Seedream 4.5
- + Highly realistic skin texture with authentic-looking scars and dirt.
- + Complex, lifelike eye rendering with excellent light catch.
- + Intricate detail on the cloth underlayer and leather straps mentioned in the prompt.
- − The armor engravings are slightly less sharp than in Model A.
- − A bit of the hair/bead detail is lost in the shallow depth of field toward the back of the head.
Verdict: Seedream 4.5 produces a more convincing and lifelike portrait with superior skin textures and realistic grit, perfectly capturing the 'battle-worn' aesthetic. While Imagen 4.0 Generate 001 excels at the sharp, intricate engravings of the armor, Seedream 4.5 offers a more cinematic composition with better depth and more realistic human features.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent adherence to the 'grid' layout request.
- + Very high-quality and diverse food photography photography.
- + Clean professional design with strong typography and geometric accents.
- − Text consists of nonsensical gibberish rather than real words.
- − The overall composition is very busy and potentially overwhelming as a real menu.
Seedream 4.5
- + Successfully includes all three requested sections (Appetizers, Pizza, Mains).
- + Visual layout is very clean and readable.
- + Modern use of color-coded borders for different categories.
- − Does not follow the 'grid' requirement, opting for a vertical list instead.
- − Repeats section titles (like 'Pizza' and 'Restaurant') as menu item names.
- − Overall design is a bit generic compared to the first option.
Verdict: Imagen 4.0 delivers a much more visually striking and complex design that perfectly follows the 'grid' requirement with high-end photography, though its text is pure gibberish. Seedream 4.5 provides a more functional-looking menu layout, but it ignored the grid instruction and has very repetitive text entries. Imagen 4.0 is the preferred choice for a professional design concept.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent typography rendering with clean, readable characters.
- + Complete adherence to the 'exploded' concept with all ingredients clearly separated.
- + Strong photorealistic textures on the lettuce and patty.
- − The starburst for the price is somewhat flat compared to the rest of the image.
- − The lighting on the burger components is a bit inconsistent with the fiery background.
Seedream 4.5
- + Dynamic sense of motion with motion-blurred ingredients.
- + Impressive fiery typography that perfectly matches the background atmosphere.
- + Excellent integration of the starburst price tag into the environment.
- − The burger is not truly 'exploded' as most components are still stacked.
- − A few artifacts around the blurred flying ingredients.
Verdict: Imagen 4.0 Generate 001 followed the 'exploded' instruction much better, providing a clear view of every individual ingredient as requested. However, Seedream 4.5 created a more visually striking and cohesive advertisement with superior integration of the fiery text and background effects, despite the burger itself being less exploded. Imagen 4.0 is preferred for its clarity and precise adherence to the layout, while Seedream 4.5 is better for overall artistic impact.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent spelling accuracy for the requested menu items.
- + Very clean and sharp text rendering.
- + Good adherence to the specific date requested.
- − The text looks like a digital font rather than realistic chalk.
- − Includes meta-text like 'Tittle', 'Menu', 'Footer', and prompt descriptions directly on the board.
- − Fails the 'elegant cursive' requirement for the title.
Seedream 4.5
- + Features a highly realistic chalk texture with dust and smudges.
- + Captures the cozy café atmosphere in the background with shallow depth of field.
- + Followed the handwriting style request much more naturally than Model A.
- − Repetitive text error on the first item ('Truffle Mushroom' and 'Risotto' listed separately with the price repeated).
- − The title is not in 'elegant cursive' as requested, though it is more stylized than the body text.
Verdict: Imagen 4.0 follows the verbatim text of the prompt and the specific items best, but it includes meta-commentary from the prompt on the board and the text looks like a digital font. Seedream 4.5 creates a much more convincing and atmospheric image with realistic chalk textures, even though it makes a minor repetitive error in the layout of the first menu item. Seedream 4.5 is preferred for its superior artistic execution and realistic 'handwritten' aesthetic.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent detail on the space suit and saddle equipment.
- + Creative integration of cosmic patterns on the horse's coat.
- + Great use of dynamic lighting and cinematic streaks of light.
- − Failed the negative constraint; the astronaut is riding the horse.
Seedream 4.5
- + Beautiful composition and vibrant color palette in the nebula.
- + Realistic texture on the horse's mane and musculature.
- + The pose is very cinematic and dynamic.
- − Failed the negative constraint; the astronaut is riding the horse.
- − The astronaut's hand gesture is somewhat awkward in a space helmet context.
Verdict: Both models failed the specific prompt instruction to place the horse on top of the astronaut, instead defaulting to the common trope of an astronaut riding a horse. Since both failed the primary logic test, the preference goes to Imagen 4.0 for its superior detail in the space suit design and the creative visual flair of stars embedded in the horse's fur.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Sharp resolution and high level of detail in the fur and clothing textures.
- + Perfect adherence to the phone-usage and bored expression instructions.
- + Excellent rendering of text on the taxi signs.
- − The capybara's paws look more like talons or claws than actual capybara webbing.
- − The composition feels slightly more like a digital render than a real photo.
Seedream 4.5
- + Very high photographic realism in lighting and depth of field.
- + The capybara's face and paws look very natural and well-integrated.
- + Beautiful bokeh effect on the Manhattan city lights.
- − Minor artifact on the businesswoman's phone hand.
- − The capybara's jacket is a bit less distinct than in the other version.
Verdict: Imagen 4.0 delivers a very clean and precise image with excellent text and detail, but it feels slightly artificial. Seedream 4.5 produces a more convincing 'photoreal' result with natural lighting and a more realistic depiction of the capybara's anatomy. Seedream 4.5 is the preferred choice for its superior atmosphere and believability as a real photograph.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent graphic design layout with a clear thorn border.
- + Accurate text rendering for all requested details.
- + Stronger gothic aesthetic in the title typography.
- − The parchment roll on the left looks slightly disconnected from the composition.
- − The art style leans more towards digital illustration than cinematic realism.
Seedream 4.5
- + Atmospheric cinematic lighting and more realistic textures on the pumpkin.
- + Clean and legible text rendering across all sections.
- + Better integration of the background elements and the moody sky.
- − Features barbed wire instead of the requested 'thorns' in the foreground.
- − The overall border is missing, with only corner spiderwebs provided.
Verdict: Imagen 4.0 followed the structural layout requests much better, providing a complete border of thorns and a cohesive parchment poster feel. Seedream 4.5 opted for a more realistic and cinematic visual style with superior lighting, but used barbed wire instead of thorns and failed to include a formal border.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent photorealistic textures on the fish and salmon roe.
- + High level of detail in the sushi rice and garnishes.
- + Well-defined lighting and shadows.
- − Completely ignored the text requirements for 'JAPAN' and 'SUSHI'.
- − Ignored the request for a solid light blue background.
- − Failed to include the flag icon.
Seedream 4.5
- + Followed all text and icon instructions perfectly with clear, bold typography.
- + Captured the 'cartoon scene' and 'miniature isometric' style requested in the prompt.
- + Correctly implemented the light blue background and small raised diorama base.
- − The textures are slightly more generic compared to the first model.
- − The garnish layout is very minimal, bordering on sparse.
Verdict: While Imagen 4.0 produces a much more realistic and appetizing image, it failed to follow almost all of the stylistic and text-based instructions. Seedream 4.5 followed the prompt comprehensively, including the specific isometric layout, colors, and text, making it the winner for prompt adherence.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
20% wins
0% ties
80% wins
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent depiction of dew sparkles on the grass
- + Each animal is clearly defined and matches the requested types perfectly
- + Vibrant, sharp colors with high clarity in the background elements
- − The lighting feels more like an illustration than a 'hyper-photorealistic' photo
- − The pose of the kitten being held by the puppy looks a bit stiff and unnatural
Seedream 4.5
- + Captures the 'hyper-photorealistic' request much more effectively through lighting and depth of field
- + Dynamic composition that conveys movement and 'playfully chasing'
- + Beautiful atmospheric god rays and bokeh effects
- − The fox kit has slightly exaggerated, 'big' cartoonish eyes compared to its realistic fur
- − The bunny is partially obscured and slightly out of focus in the background
Verdict: Seedream 4.5 is the winner because it better captures the 'hyper-photorealistic' and 'playful' essence of the prompt through dynamic lighting and realistic textures. While Imagen 4.0 provides a very clear and colorful scene, it looks more like a high-quality storybook illustration rather than a photograph, whereas Seedream 4.5 creates a more immersive and atmospheric image.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent text legibility and accuracy
- + Clean and modern vector emblem style
- + Well-balanced minimalist composition
- − Banner element is a very simple shape rather than a classic banner
- − Typography feels a bit too modern for a 'vintage' request
Seedream 4.5
- + Stronger vintage aesthetic with serif typography and hatching details
- + Traditional banner design for the 'Est. 1720' text
- + Good use of subtle paper texture on the background
- − The accent mark on 'Caffè' is slightly disconnected and oddly shaped
- − Text alignment on the curved path is slightly irregular
Verdict: Both models followed the prompt well, but they offer different interpretations of 'vintage.' Imagen 4.0 provides a clean, modern-minimalist take with perfect text rendering, while Seedream 4.5 captures the 'vintage' and 'retro' keywords much more effectively through its use of serif fonts, a classic banner, and woodcut-style shading. Seedream 4.5 is the preferred choice for matching the specific historical and artistic style requested.
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.”
Imagen 4.0 Generate 001
Seedream 4.5
AI Judge Analysis
Imagen 4.0 Generate 001
- + Excellent typography and header placement.
- + Strong architectural layout showing the Saturn V rocket alongside the steps.
- + Captures the NASA-inspired color palette perfectly.
- − Failed to include steps 5 and 6 (Descent and Landing).
- − The sequence of the infographic is logically confusing and repetitive.
- − Iconography for the Earth and Moon is inconsistent across different steps.
Seedream 4.5
- + Successfully included all 6 requested steps in a clear chronological timeline.
- + Accurate and crisp flat-vector iconography for the Lunar Module and Saturn V.
- + Creative inclusion of the crew names and landing site 'Tranquility' as extra details.
- − Step 5 (Descent) is represented by a generic satellite icon instead of a descending lunar module.
- − The 'Earth Orbit' illustration is physically incorrect with the ring passing through the planet.
- − Vertical layout of labeling takes up significant white space at the bottom.
Verdict: Seedream 4.5 is the clear winner as it followed the complex multi-step instructions and maintained a logical chronological order for the infographic. While Imagen 4.0 has superior visual polish and professional typography, it failed several core prompt requirements by omitting the final two steps of the mission.
Next steps
Explore each model
ByteDance's latest image generation model unifying text-to-image and image editing in a single architecture, with improved text rendering and 30-40% faster generation than v4.0