HiDream AI's 17B parameter text-to-image model using sparse diffusion transformer with mixture of experts, achieving state-of-the-art image generation quality with strong prompt following
Settled by community votes across 5 shared challenges, with an AI judge weighing in on each.
HiDream I1 Full
13.7
arena score
#60 of 62 in Text-to-Image
Skill signature
· Text-to-Image
Imagen 3.0 Generate 002
21.0
arena score
#36 of 62 in Text-to-Image
Vote tally
Where the votes landed
HiDream I1 Full
0%
win rate
Ties
0%
Imagen 3.0 Generate 002
0%
win rate
Shared challenges
5
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.”
HiDream I1 Full
Imagen 3.0 Generate 002
AI Judge Analysis
HiDream I1 Full
- + Excellent text legibility and alignment
- + Very clean, minimalist aesthetic
- + High-quality, appetizing food photography
- − Lacks variety in the food photos, showing only pizzas
- − Layout is a bit sparse for a full menu
Imagen 3.0 Generate 002
- + Strong adherence to the grid layout request
- + Greater variety of food types representing different sections
- + Sophisticated professional layout with secondary details like URLs
- − Most of the smaller text is illegible gibberish
- − The grid feels a bit cramped and repetitive
Verdict: HiDream I1 Full creates a very clean and usable design with legible text, but fails to provide variety in the food photos as it only shows pizzas. Imagen 3.0 Generate 002 follows the 'grid' instruction much more effectively and provides a more complex, professional-looking design, though the body text is not readable.
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.”
HiDream I1 Full
Imagen 3.0 Generate 002
AI Judge Analysis
HiDream I1 Full
- + Successfully rendered the date and a title similar to the prompt
- + Good chalk illustration of a coffee cup
- + Realistic chalkboard frame and background blur
- − The main text is entirely nonsensical gibberish
- − The 'handwriting' looks more like a digital font than natural chalk variations
- − Failed to include any of the specifically requested menu items
Imagen 3.0 Generate 002
- + Excellent adherence to the specific menu items requested
- + Very realistic chalk texture with smudges and natural variations
- + Mostly accurate spelling of complex words like 'Truffle Mushroom Risotto'
- − Includes some repetition and minor spelling distortions at the bottom
- − The title is blocky rather than the 'elegant cursive' requested
Verdict: Imagen 3.0 Generate 002 followed the prompt's instructions for specific menu items and prices much more accurately than HiDream I1 Full, which produced high-quality visuals but completely failed on the text content. Imagen 3.0 Generate 002 also captured the requested chalk texture and natural handwriting variations far more convincingly.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
HiDream I1 Full
Imagen 3.0 Generate 002
AI Judge Analysis
HiDream I1 Full
- + Excellent dynamic lighting with a high-contrast cinematic feel.
- + Great planet-side perspective that adds a sense of scale and movement.
- + Clear, sharp rendering of both the astronaut's suit and the horse's anatomy.
- − The horse's back left leg is anatomically inconsistent with its position.
- − The reins disappear into the astronaut's hand rather than being gripped realistically.
Imagen 3.0 Generate 002
- + Beautiful cosmic background with a detailed nebula effect.
- + Technically superior rendering of textures, especially the horse's mane and the space suit's fabric.
- + Very coherent composition with realistic placement of the rider on the saddle.
- − The overall lighting is a bit flat compared to the dramatic shadows in Model A.
- − Slightly less 'cinematic' in terms of angle and perspective.
Verdict: Both models failed the negative constraint to put the 'horse on top' of the astronaut, instead interpreting the prompt as an astronaut riding a horse. Imagen 3.0 Generate 002 is the winner due to its superior texture handling, more coherent anatomy, and beautiful nebula background, whereas HiDream I1 Full has more dramatic lighting but suffers from anatomical glitches in the horse's legs.
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.”
HiDream I1 Full
Imagen 3.0 Generate 002
AI Judge Analysis
HiDream I1 Full
- + Excellent high-clarity text rendering for both 'JAPAN' and 'SUSHI'.
- + Sophisticated 3D rendering with convincing subsurface scattering and PBR materials.
- + Clean, professional-looking diorama base that fits the isometric theme.
- − Failed to include the requested flag icon.
- − The text is centered but overlaps slightly with the blue background's gradient texture.
Imagen 3.0 Generate 002
- + Included all requested elements, including the small flag icon next to the text.
- + Good adherence to the 45-degree isometric perspective and diorama base requirement.
- + Clean, soft cartoon aesthetic that matches the 'miniature' prompt description.
- − The text is aligned to the top-left instead of top-center.
- − The overall textures are a bit flatter and less 'refined' or 'realistic PBR' compared to the other model.
Verdict: HiDream I1 Full produces a much higher quality 3D render with superior materials and lighting, but it fails to include the flag icon. Imagen 3.0 Generate 002 is more accurate to the specific details of the prompt (including the flag) but places the text in the wrong location and features less impressive material work. HiDream I1 Full is the likely winner for its professional visual fidelity and much clearer text rendering.
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.”
HiDream I1 Full
Imagen 3.0 Generate 002
AI judge analysis unavailable for this challenge.
Next steps
Explore each model
Google's Imagen 3.0 text-to-image generation model, producing high-quality images with improved detail and lighting