Gemini 3.1 Flash with image generation capabilities. High-efficiency image generation model with support for text rendering, reference images, search grounding, and thinking mode. The efficient counterpart to Gemini 3 Pro Image.
Settled by community votes across 6 shared challenges, with an AI judge weighing in on each.
Nano Banana 2
28.4
arena score
#1 of 62 in Text-to-Image
Best Text-to-Image right now
Top 2 in Image Editing
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
Nano Banana 2
0%
win rate
Ties
0%
Imagen 3.0 Generate 002
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.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent text rendering with highly legible and accurate food names and descriptions.
- + Outstanding layout that perfectly categorizes appetizers, pizza, and mains into a structured grid.
- + Very high-quality food photography that aligns perfectly with the adjacent text descriptions.
- − Minor spelling errors in small descriptive text like 'somon' instead of salmon.
- − Slightly crowded footer text at the bottom edge.
Imagen 3.0 Generate 002
- + Good use of white space and a clean, square-grid aesthetic.
- + Consistent lighting across the food photography elements.
- − Text is mostly gibberish or heavily distorted, failing to provide a functional menu.
- − The layout is repetitive and confusing, with section headers like 'PIZZA' appearing in multiple disconnected boxes.
- − Poor adherence to the section requirements, as 'Appetizers' is misspelled and poorly integrated.
Verdict: Nano Banana 2 is the clear winner as it produces a fully functional, professional-grade restaurant menu with readable text and logical categorization. In contrast, Imagen 3.0 Generate 002 fails to render legible English text and creates a confusing, repetitive layout that lacks the clarity required for a menu design.
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.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent typography rendering with the requested fiery, glowing effect.
- + Highly photorealistic textures on the patty, bun, and vegetables.
- + Dynamic composition with a strong sense of suspended motion and splash effects.
- − The price tag starburst is a bit large, slightly crowding the central image.
Imagen 3.0 Generate 002
- + Clean, professional-looking food product layout.
- + Vibrant colors and clear separation of burger layers.
- − Fails to render the text with the requested 'fiery effect' as convincingly as the other model.
- − Includes nonsensical gibberish text inside the price starburst.
- − The lighting on the burger components feels a bit flatter and less integrated with the background.
Verdict: Nano Banana 2 is the superior image as it perfectly follows the typographic instructions, including the fiery glow effect on all text elements. While Imagen 3.0 Generate 002 creates a clean layout, it suffers from significant AI text artifacts in the starburst and lacks the dynamic, photorealistic energy found in Nano Banana 2.
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.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent text accuracy with no spelling errors.
- + Authentic chalk texture and handwriting variations.
- + Great composition that captures the requested café atmosphere.
- − The 'cursive' for the title is a bit more of a stylized print than true elegant cursive.
Imagen 3.0 Generate 002
- + Good blackboard texture and framing.
- + Uses different colors for the date as a creative touch.
- − Significant spelling errors throughout the menu items such as 'Risotutto' and 'Choouip'.
- − The text looks more like a digital font than natural chalk handwriting.
- − The layout is cluttered with repeated/mangled lines of text.
Verdict: Nano Banana 2 followed the prompt with near-perfect accuracy, providing clear and correctly spelled text in a realistic café setting. In contrast, Imagen 3.0 Generate 002 struggled with the complex text instructions, resulting in several typos, repeated lines, and a less convincing handwriting style.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent adherence to the 'horse on top' spatial instruction
- + Vibrant cinematic lighting and textures
- + Creative interpretation of the surreal requirement
- − The astronaut's leg position relative to the saddle is physically nonsensical
- − Some anatomical warping in the horse's neck and legs
Imagen 3.0 Generate 002
- + Clean, high-quality rendering of the subjects
- + Well-composed and balanced frame
- + Detailed texture on the astronaut suit and horse's coat
- − Completely failed the negative constraint to have the horse on top
- − Lacks the 'surreal' quality requested, looking like a standard sci-fi image
Verdict: Nano Banana 2 was the only model to successfully follow the complex spatial instruction of putting the horse on top of the astronaut, resulting in a truly surreal image. Imagen 3.0 Generate 002 ignored the specific 'horse on top, not vice versa' instruction, producing a generic (though high-quality) image of an astronaut riding a horse.
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.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent text rendering with clean typography.
- + High-quality PBR textures for the wood, rice, and fish.
- + Great complexity in terms of sushi variety and diorama details like moss.
- − The text is placed over the flag rather than having the flag as a separate small icon alongside it as requested.
- − The diorama base is quite busy compared to the 'minimal' request.
Imagen 3.0 Generate 002
- + Perfectly captures the 'cartoon' and 'miniature' aesthetic with soft, rounded textures.
- + Correct architectural isometric 45-degree angle.
- + Clean and minimal diorama base exactly as described.
- − The text and icon are placed in the top-left rather than top-center as requested.
- − The 'SUSHI' text is slightly darker and less prominent than the 'JAPAN' text.
Verdict: Nano Banana 2 produces a more realistic and visually impressive sushi display, but Imagen 3.0 Generate 002 follows the stylistic 'cartoon' and 'isometric' instructions much more accurately. While Nano Banana 2 has better text placement, Imagen 3.0 Generate 002 wins on the composition and the specific requested 3D miniature art style.
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.”
Nano Banana 2
Imagen 3.0 Generate 002
AI Judge Analysis
Nano Banana 2
- + Excellent depiction of god rays and sunrise lighting consistent with the prompt.
- + Each of the four requested animals is clearly distinct and correctly identified.
- + Dynamic composition that suggests the 'chasing' and 'tumbling' action.
- − The fox kit has slightly unnatural front legs while running.
- − Butterflies feel a bit like flat stickers layered onto the scene.
Imagen 3.0 Generate 002
- + Superb fur texture and detail, particularly on the golden retriever and kitten.
- + Captures the 'tumbling' and 'wholesome' vibe very effectively with the bunny on its back.
- + Beautiful bokeh and dewdrop effects in the grass.
- − The fox and kitten are somewhat blended together in the mid-ground, making them harder to distinguish.
- − The 'chasing' aspect of the prompt is less apparent than the 'tumbling' aspect.
Verdict: Both models performed exceptionally well on this complex prompt. Nano Banana 2 features a more dynamic composition that captures the 'running and chasing' movement with clear distinction between all four animals, while Imagen 3.0 Generate 002 excels in hyper-realistic fur texture and the 'tumbling' aspect of the request. Nano Banana 2 is the slight winner for better adherence to the specific 'chasing butterflies' movement and more prominent god rays.
Next steps
Explore each model
Google's Imagen 3.0 text-to-image generation model, producing high-quality images with improved detail and lighting