GPT Image 2 OpenAI vs Imagen 4.0 Ultra Generate 001 Google

GPT Image 2

• + Excellent adherence to the 'partially visible through the glass' instruction.
• + Highly realistic texture on the red book and the glass framing.
• + Natural and soft window lighting that feels physically accurate.

• − The glass cube looks more like a hollow glass box with thick frames rather than a solid or thin-walled cube.

Imagen 4.0 Ultra Generate 001

• + Beautifully rendered light rays and shadows on the wooden table.
• + Clean, solid glass material with realistic refraction indices.
• + Detailed book spine with legible, high-quality text.

• − The blue sphere appears to be floating mid-air inside the cube without physical support.
• − The plant is entirely behind the cube and its visibility through the glass is less natural than in model A.

GPT Image 2

• + Excellent depiction of motion blur from passing cars as requested.
• + Realistic candid framing with interesting foreground elements like the sign.
• + Very natural skin textures and believable lighting on the wet pavement.

• − The red bicycle frame looks a bit thin and structurally confusing near the seat post.

Imagen 4.0 Ultra Generate 001

• + Beautiful rain droplet effects on the character's jacket.
• + Strong cinematic composition with clear shallow depth of field.
• + High-quality skin texture and detail on the subject's hands.

• − Missed the 'motion blur from passing cars' requirement, as the car is static and sharp.
• − The bicycle is floating against the wall without a kickstand or leaning support.

GPT Image 2

• + Natural, cinematic lighting with a shallow depth of field.
• + Photorealistic skin textures and lifelike eyes.
• + Highly detailed and believable weathered plate armor.

• − The braids and beads are a bit subtle compared to the request.
• − The 'battle-worn' aspect is mostly dirt rather than visible scars.

Imagen 4.0 Ultra Generate 001

• + Excellent adherence to specific details like the beads in the hair and prominent scars.
• + Dynamic composition with visible sparks and a torch.
• + Clear, high-contrast engraving on the armor.

• − The character's skin and hair have a slightly plastic, CGI look.
• − The torch on the shoulder appears to be floating or awkwardly attached.
• − The symmetry and lighting feel less natural and more staged.

GPT Image 2

• + Exceptional text rendering with perfect spelling and coherent menu item descriptions.
• + Highly professional layout with sophisticated graphic design elements and branding.
• + Food photography is high-quality and consistent with the types of food described.

• − The font hierarchy is slightly dense in the descriptions, but still very legible.

Imagen 4.0 Ultra Generate 001

• + Successfully follows the grid layout request for food photos.
• + Features a clean white background as requested.

• − Text is comprised of gibberish and artifacts rather than readable English.
• − The layout is imbalanced with excessive empty space and illogical section groupings.
• − Food photos are of lower visual quality and lack consistent plate presentation.

GPT Image 2

• + Excellent photorealistic texture on the burger bun, meat, and vegetables.
• + Dynamic and messy 'exploded' effect with realistic sauce splashes and flying embers.
• + Strong adherence to the 'fiery glowing' text effect for all requested elements.

• − The composition feels a bit crowded with large text overlapping the sparks.

Imagen 4.0 Ultra Generate 001

• + Clean, symmetrical composition that works well for a traditional advertisement.
• + Accurate text rendering and placement of the price starburst.
• + Good lighting on the food components.

• − The food assets look slightly more artificial/plastic compared to the other model.
• − The 'exploded' effect is very static and lacks the requested sense of motion.
• − The background is a simple swirl rather than a detailed fiery environment with embers.

GPT Image 2

• + Excellent chalk texture throughout all characters
• + Perfect adherence to handwriting request with natural slant and variable line thickness
• + Realistic café background and lighting

• − Minor cursive connection error in the word 'TODAY'S'

Imagen 4.0 Ultra Generate 001

• + Perfectly legible and accurate text rendering
• + Good use of compositional wipes on the chalkboard
• + Clear, high-contrast text

• − Text looks like a digital handwritten-style font rather than organic chalk
• − Edges of letters are too smooth and lack grainy chalk texture
• − The 'cursive' request for the title was not fully met

GPT Image 2

• + Perfectly follows the specific instruction of 'horse on top'
• + Highly detailed textures on the spacesuit and horse fur
• + Convincing cinematic lighting and lunar environment

• − The horse's front legs and harness anatomy are a bit nonsensical near the astronaut's shoulders
• − The astronaut's hands have slightly irregular finger counts

Imagen 4.0 Ultra Generate 001

• + Clean, illustrative style with vibrant colors
• + Creative use of space gear for the horse
• + Good composition with asteroids and ringed planet

• − Completely failed the negative constraint/specific instruction by putting the astronaut on top
• − Lacks the 'cinematic' realism requested
• − Perspective on the horse's legs is slightly distorted

GPT Image 2

• + Features a highly photorealistic texture on the capybara's fur and the surrounding taxi interior.
• + Captures the 'bored' expression of the passenger perfectly, contributing to the deadpan humor requested.
• + Atmospheric lighting through the window feels authentic to a rainy night in NYC.

• − The passenger is positioned slightly further away than ideal for a tight interior shot.
• − The capybara's hands are rendered somewhat like human fingers wrapped in fur rather than paws.

Imagen 4.0 Ultra Generate 001

• + Excellent composition that shows both the driver and the passenger clearly in one frame.
• + Very accurate adherence to the 'both paws on the wheel' instruction with distinct, sharp claw details.
• + Clarity and lighting are very strong, giving it a polished cinematic look.

• − The passenger looks slightly distressed or anxious rather than 'bored'.
• − The overall image has a slightly digital, CGI-like finish compared to the photorealism of the other image.
• − The capybara's head shape is a bit elongated/distorted.

GPT Image 2

• + Perfectly captures the 'vintage gothic' and 'dark parchment' aesthetic requested.
• + Excellent text rendering with no spelling errors across all three sections.
• + Highly detailed composition featuring specific NYC references like high-rise silhouettes and arches.

• − The dark color palette makes the border detail slightly harder to distinguish than model_b.

Imagen 4.0 Ultra Generate 001

• + Clean, legible layout with accurate text rendering.
• + Strong color contrast with vibrant orange and blue tones.
• + Faithfully includes all required elements like the scroll and thorn border.

• − Lacks the 'vintage' and 'dark parchment' feel, appearing more like a modern digital illustration.
• − Thorn border looks somewhat artificial and flat compared to the requested cinematic gothic style.

GPT Image 2

• + Excellent 3D text rendering with shadow and depth
• + High richness of detail in the sushi textures and diorama base
• + Creative inclusion of a stone lantern and diverse nigiri types

• − The scene is a bit more crowded than the requested 'minimal garnish' design

Imagen 4.0 Ultra Generate 001

• + Perfect adherence to the 45-degree isometric camera angle
• + Clean, minimal aesthetic with a clear distinction between the base and plate
• + Soft, clay-like cartoon textures that match the prompt perfectly

• − The text layout is a bit scattered and lacks the 3D 'pop' requested for a cartoon scene
• − Composition is a bit floating and detached compared to a traditional diorama

GPT Image 2

• + Excellent photorealistic texture on the fur and grass.
• + Natural lighting with realistic 'god rays' and backlighting.
• + High degree of anatomical accuracy and dynamic posing.

• − The fox kit has black front legs that look slightly like solid sleeves/socks.

Imagen 4.0 Ultra Generate 001

• + Strong composition that captures the 'tumbling' and 'chasing' aspect of the prompt.
• + Clear depiction of dew sparkles on the flowers and grass.
• + Whimiscal, vibrant colors that fit a 'joyful' vibe.

• − The style leans more toward a high-end digital illustration or '3D render' than the requested 'hyper-photorealistic' scene.
• − Anatomical issues on the cat's front right paw and the fox's body connection.

GPT Image 2

• + Excellent typography with a sophisticated vintage serif font.
• + Beautifully detailed etching style on the cloche and banner.
• + Strong composition with a professional frame that enhances the logo's impact.

• − The design is quite ornate, leaning more toward 'vintage' than the requested 'minimalist' style.

Imagen 4.0 Ultra Generate 001

• + Successfully captures the requested minimalist vector aesthetic.
• + Clean and legible layout with accurate text rendering.
• + Good adherence to the brown and cream color scheme.

• − The steam element is very simple and lacks the elegant 'vintage' feel found in the other model.
• − The typography is basic and less characteristic of a premium restaurant logo.

GPT Image 2

• + Excellent text rendering with accurate labels and astronaut names.
• + Perfect adherence to all six requested steps with appropriate icons for each.
• + Clean, professional composition that follows the NASA-inspired color palette perfectly.

• − The illustrations use slightly more 3D shading than the 'flat-vector' request.
• − The iconic EAGLE landing patch in the corner includes a stylized eagle that is slightly unrefined.

Imagen 4.0 Ultra Generate 001

• + Successfully captures a flat, minimalist vector aesthetic.
• + Adheres well to the requested color palette.

• − Fails significantly on text rendering, producing gibberish for most labels.
• − Does not follow the 6-step chronological sequence requested in the prompt.
• − The layout is cluttered and the iconography is abstract and confusing.