Google's Imagen 3.0 text-to-image generation model, producing high-quality images with improved detail and lighting
Settled by community votes across 6 shared challenges, with an AI judge weighing in on each.
Imagen 3.0 Generate 002
21.0
arena score
#36 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Stable Diffusion 3.5 Large
22.9
arena score
#29 of 62 in Text-to-Image
Vote tally
Where the votes landed
Imagen 3.0 Generate 002
0%
win rate
Ties
0%
Stable Diffusion 3.5 Large
0%
win rate
Shared challenges
6
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
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 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Excellent adherence to the grid layout requested
- + Clean, professional typography that is highly legible
- + Consistent lighting and high visual quality in food photography
- − Minor spelling errors in category headers like 'APPETIEES'
- − The text blocks contain repetitive gibberish characters
Stable Diffusion 3.5 Large
- + Strong 'modern minimalist' aesthetic with bold central typography
- + Vibrant, high-contrast food photos
- + Creative use of a vertical center column layout
- − Layout is cut off at the edges, failing the professional design requirement
- − Significant spelling errors in every single header
- − Grid structure is less orderly than Model A
Verdict: Imagen 3.0 Generate 002 is the superior choice because it provides a complete, usable layout that follows the grid prompt precisely while maintaining high legibility. Stable Diffusion 3.5 Large has a compelling artistic style but fails on the basics of professional design, with cut-off borders and more severe text hallucinations.
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 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Perfect adherence to all text requirements including 'MAGIC BURGER', 'LIMITED TIME ONLY', and the '€6.99' price in a starburst.
- + Excellent 'exploded burger' effect with clear separation and suspension of all requested components.
- + Vibrant, commercial-grade lighting and background that fits the 'fiery' and 'glowing' theme.
- − The secondary text inside the starburst badge is gibberish.
- − The sauce droplets have a slightly plastic, CGI-like texture compared to the photorealistic ingredients.
Stable Diffusion 3.5 Large
- + Highly realistic texture on the beef patties and the charred surface of the bun.
- + Dramatic lighting with intense flame effects and glowing embers at the base.
- − Completely failed to include any of the requested text elements (titles, price, or starburst).
- − Did not follow the 'exploded burger' instruction; the components are stacked rather than suspended in mid-air.
- − The bottom of the bun appears to be melting into the fire, which is a bit messy.
Verdict: Imagen 3.0 successfully followed every part of the complex prompt, including delicate text rendering and the specific 'exploded' layout required for the advertisement. In contrast, Stable Diffusion 3.5 Large failed to include any text or the exploded effect, resulting in a standard burger image instead of the requested ad. While Stable Diffusion 3.5 Large has impressive texture work, Imagen 3.0 is the clear winner for its superior prompt adherence and composition.
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 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Excellent chalk texture and realistic handwriting variation
- + Correctly included the date 2026 requested in the prompt
- + Clean wood frame and high-quality legible text rendering
- − Several spelling errors like 'Risotito' and 'Choouip'
- − Redundant text repetition at the bottom of the board
Stable Diffusion 3.5 Large
- + Beautiful environmental context showing a cozy café interior
- + Aesthetically pleasing layout with illustrative framing elements on the board
- − Text is highly garbled and incoherent across all menu items
- − Failed to use the correct year (2024 instead of 2026)
- − Significant spelling error in the title ('TODAAY')
Verdict: Imagen 3.0 Generate 002 is the clear winner for its superior text rendering, chalk texture realism, and adherence to the specific date requested. While Stable Diffusion 3.5 Large provides much better context for the 'cozy café' part of the prompt, the text itself is nearly unreadable and contains far more errors than Imagen's output.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
Imagen 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Excellent anatomical rendering of the horse and astronaut
- + Clean, cinematic lighting with a deep nebula background
- + High level of detail in the astronaut's suit textures
- − The horse has protective wraps on its legs that appear slightly out of place
- − Failed to follow the surreal instruction for 'horse on top'
Stable Diffusion 3.5 Large
- + Dynamic composition with a sense of motion through cosmic clouds
- + Atmospheric lighting reflecting off the space suit
- + Intricately detailed saddle and harness equipment
- − Anatomy of the horse's neck and head is awkward and elongated
- − Completely ignored the spatial logic instruction for 'horse on top'
Verdict: Both models failed to correctly interpret the counter-intuitive instruction for the horse to be 'on top' of the astronaut, instead providing standard astronaut-on-horse images. Imagen 3.0 is the superior choice because it offers much cleaner anatomical rendering and a more polished, cinematic aesthetic compared to the distorted proportions of the horse in Stable Diffusion 3.5 Large.
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 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Excellent adherence to the 3D cartoon style and soft refined textures.
- + Clean and professional text rendering with a minimalist layout.
- + Perfect isometric perspective and composition.
- − Text is placed in the top-left rather than top-center as requested.
Stable Diffusion 3.5 Large
- + High detail in the sushi textures and rice grains.
- + Includes more variety of sushi types.
- − Failed the 'cartoon' style requirement, producing a semi-realistic render instead.
- − Failed the 'top-center' text placement, including a physical sign and flag instead of a clean graphic overlay.
- − Composition feels cluttered compared to the requested 'minimal' look.
Verdict: Imagen 3.0 Generate 002 significantly outmatched the other model by strictly following the requested 3D cartoon art style and clean isometric presentation. While Stable Diffusion 3.5 Large provided more texture detail, it failed the stylistic prompt and created a messy composition with physical objects representing the requested text overlay.
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 3.0 Generate 002
Stable Diffusion 3.5 Large
AI Judge Analysis
Imagen 3.0 Generate 002
- + Excellent fur texture rendering and sharpness
- + Highly realistic anatomical features for each animal
- + Coherent composition with clear lighting and dew effects
- − The animals are mostly sitting/lying down rather than actively 'chasing' and 'tumbling'
- − The god rays are a bit subtle compared to the prompt's request
Stable Diffusion 3.5 Large
- + Successfully captures the action of 'chasing' and running
- + Bright, joyful atmosphere with prominent golden light and bokeh
- + Expressive, cheerful facial expressions on the animals
- − Anatomical issues, particularly the fox's face which looks slightly distorted
- − Lower overall resolution and detail in the fur compared to Model A
- − The kitten is missing characteristic tabby markings
Verdict: Imagen 3.0 Generate 002 produces a significantly higher quality image in terms of photorealism and fine detail, particularly in the fur and facial structures. Stable Diffusion 3.5 Large does a better job of capturing the specific motion of 'chasing' requested in the prompt, but it suffers from some anatomical clipping and less realistic textures.
Next steps
Explore each model
Stability AI's 8.1-billion parameter Multimodal Diffusion Transformer (MMDiT) text-to-image model featuring improved image quality, typography, complex prompt understanding, and resource-efficiency