Grok Imagine Image xAI vs OmniGen v2 VectorSpaceLab

Grok Imagine Image

• + Excellent photorealistic texture on the wooden table and book.
• + Realistic glass refractions and caustics on the table surface.
• + Accurately places the plant behind the glass with realistic distortion.

• − The blue sphere is levitating unnaturally in the center of the cube.
• − The cube is more of a hollow rectangular prism than a solid geometric cube.

OmniGen v2

• + Solid cube geometry with a logical placement of the blue sphere on the bottom surface.
• + Clean, vibrant colors and clear adherence to all prompt elements.
• + Strong lighting consistency and soft shadows.

• − The texture of the wooden table is a bit flat and looks more like laminate.
• − The plant behind the glass lacks the realistic refractive distortion seen in the other model.

Grok Imagine Image

• + Excellent adherence to the 'imperfect framing' and 'candid' street photography style.
• + Realistic motion blur from the passing vehicle and natural film-like grain.
• + Highly authentic bicycle details and believable lighting.

• − The man's face is mostly hidden, though this fits the 'candid' instruction.
• − Slightly less clarity on the man's hands.

OmniGen v2

• + Clear facial details and identifiable features.
• + Nice reflections on the wet pavement.

• − Fails on the 'motion blur from passing cars' instruction as the car is stationary.
• − The rain effect looks like a static overlay rather than a natural environmental element.
• − The bicycle geometry is incorrect, particularly where the frame meets the rear wheel and pedal area.

Grok Imagine Image

• + Extremely detailed and intricate ornate engraving on the plate armor
• + More realistic skin texture and lifelike eyes
• + Successful execution of the cloth underlayer and leather straps with high-resolution texture

• − The sparks look slightly like static dots in some areas

OmniGen v2

• + Good interpretation of hair braids and beads
• + Clean composition with clear warm lighting focus

• − Armor engravings are much simpler and lack the 'ornate' quality of the prompt
• − Leather straps and cloth layers are missing or very basic in detail
• − The dirt on the face looks specifically like painted dots rather than 'battle-worn' grime

Grok Imagine Image

• + Excellent typography with mostly legible section headers and item names
• + Clean, professional layout that closely mimics a real-world restaurant menu
• + The food photography is integrated naturally with the text, creating high visual appeal

• − Contains several repeating items like 'Steak Frites' and 'Grilled Salmon'
• − Some minor spelling errors in smaller text and headers (e.g., 'Veggie Deasiy')

OmniGen v2

• + Strong use of vibrant color blocks as accents
• + Clear grid structure for the top section photography

• − Severe spelling errors in headers (e.g., 'Apptetizes', 'Restaurated Ments')
• − The layout feels more like a flyer or a rough template than a functional professional menu
• − Lower resolution/clarity in the food imagery compared to the other model

Grok Imagine Image

• + Excellent adherence to the 'exploded burger' requirement with dynamic component suspension.
• + Text is perfectly rendered, including the specific currency symbol and fiery effect.
• + Photorealistic texture on the meat patty, lettuce, and bun.

• − The starburst element for the price is a bit busy compared to the rest of the professional composition.

OmniGen v2

• + Bold, clean graphic design that feels like a professional fast-food advertisement.
• + Good integration of the logo with the fiery background glow.

• − Completely failed the 'exploded burger' request, showing a static, assembled burger.
• − Text is cut off on the left side ('LIMITED TIME ONLY').
• − Missing the Euro symbol (€) requested in the price starburst.

Grok Imagine Image

• + Excellent text rendering with perfect spelling and grammar.
• + Realistic chalk texture with convincing dust and smudges on the board.
• + Consistent handwriting style that maintains a natural, non-digital look.

• − The title is in print-style handwriting rather than the requested 'elegant cursive'.

OmniGen v2

• + Good contrast between the chalkboard and the frame.
• + Layout follows a traditional menu structure.

• − Significant spelling errors throughout the text (e.g., 'SPECALS', 'Lemont', 'Octapus').
• − Text becomes unintelligible gibberish in the middle sections.
• − The handwriting looks more like a digital brush than natural chalk on a textured surface.

Grok Imagine Image

• + Successfully follows the logical constraint of the horse being 'on top' through a surreal interpretation.
• + Excellent use of lighting, color, and cinematic nebula effects.
• + High level of detail in the astronaut's suit and the horse's anatomy.

• − The horse has an extra leg merging into its front right area.
• − The relationship between the horse's hoof and the astronaut's hand is slightly ambiguous.

OmniGen v2

• + Clean, simple composition with clear subjects.
• + Solid rendering of the astronaut's gear and the horse's coat.

• − Completely fails the negative constraint of 'horse on top, not vice versa'.
• − Flat, less detailed background compared to the 'cinematic' request.
• − Anatomical errors in the horse's legs and hooves.

Grok Imagine Image

• + Excellent photorealism with realistic lighting and textures.
• + Accurately follows the prompt by placing the capybara in the driver's seat and the passenger in the back.
• + Capture the requested 'bored' expression of the passenger perfectly.

• − The capybara's paws are rendered more like claws/fingers than natural capybara anatomy.

OmniGen v2

• + Features a sharp, high-contrast image.
• + The capybara's face is well-rendered with clear textures.

• − Significant failure in composition: the passenger is sitting in the driver's seat with the capybara on her lap or merged through her.
• − Human hands are shown steering the wheel instead of the capybara's paws.
• − The passenger is in the front seat instead of the requested back seat.

Grok Imagine Image

• + Excellent text rendering with perfect spelling and placement.
• + Strong adherence to all visual elements including thorns, webs, and scroll banner.
• + Sophisticated cinematic lighting and authentic vintage parchment texture.

• − The transition between the top thorns and the moon feels slightly cluttered.

OmniGen v2

• + Bold, high-contrast color palette with a vibrant glow.
• + Clear and legible title text.

• − Failed to render the event details accurately, with numerous misspellings and garbled text.
• − The scroll banner contains illegible gibberish.
• − The composition feels flatter and more like a modern digital graphic than a vintage gothic invitation.

Grok Imagine Image

• + Excellent text rendering and placement that perfectly follows the prompt.
• + Realistic 3D material feel with good wood grain and subsurface scattering on the fish.
• + Captures the Japanese flag correctly above the text.

• − The sushi rice grains look slightly like bubbles or pearls rather than traditional rice.

OmniGen v2

• + Clean isometric diorama base with distinct layers.
• + Good 3D drop shadow effect on the typography.

• − Fails to include the Japanese flag correctly, showing a generic yellow/red flag instead.
• − The sushi design is illogical, featuring a 'tail' on one piece that doesn't match standard sushi types.
• − Less variety in the sushi types compared to the other model.

Grok Imagine Image

• + Successfully included all four requested animals: puppy, kitten, bunny, and fox.
• + Dynamic and energetic composition showing the animals actually tumbling and chasing.
• + Better rendering of textures with more realistic lighting and dew sparkles.

• − The fox kit has somewhat stylized, oversized features relative to the others.
• − The background is quite blurry, though it fits the bokeh request.

OmniGen v2

• + Clearly includes the butterflies mentioned in the prompt.
• + Bright, vibrant colors that fit a 'wholesome' vibe.

• − Failed to include the baby bunny, showing only three animals.
• − The image has a very flat, digital cartoon-like aesthetic rather than 'hyper-photorealistic'.
• − The pose is static and misses the 'tumbling together' action requested.

Grok Imagine Image

• + Excellently captures the painterly, watercolor-textured background typical of Ghibli films
• + Maintains the specific facial expressions and character features of the original people
• + Preserves the structure and architectural detail of the background faithfully

OmniGen v2

• + Successfully translates the scene into a clean anime art style
• + Matches the character poses and clothing colors accurately

• − Faces look generic and have lost the distinct 'distracted' and 'indignant' expressions of the original meme
• − Colors are too saturated and flat, failing the 'soft pastel' and 'hand-painted' requirements
• − Lighting is flat and lacks the requested warm, nostalgic mood

Grok Imagine Image

• + Excellent prompt adherence with numerous flying leaves throughout the scene.
• + Strong preservation of the original subject's face, clothing, and background details.
• + Dynamic hair and canine ear motion that feels natural to the source.

• − The sheer number of leaves can feel a bit cluttered and overlayed.
• − Slight loss in texture sharpness compared to the original image.

OmniGen v2

• + Natural and elegant hair motion blowing to the left.
• + Clean composition with few distractions.

• − Failed to include sufficient 'flying leaves,' with only a few stylized orange shapes present.
• − Significant changes to the subject's face and the dog's appearance, moving away from being a faithful edit.
• − The environment textures (grass and flowers) have been smoothed out into a more illustrative style.

Grok Imagine Image

• + Excellent typography with correct spelling and accent marks
• + Sophisticated vector style with subtle grain texture
• + Includes a creative 'caffe' cup hidden within the cloche design

• − Repeats the 'Est. 1720' text twice which creates layout redundancy

OmniGen v2

• + Follows the 'banner' requirement for the typography effectively
• + Clean minimalist vector lines
• + Good use of warm cream and brown tones

• − Major spelling errors in the brand name ('CAFFFLORIN')
• − Steam effect is very thin and lacks the requested vintage feel

Grok Imagine Image

• + Successfully included all six requested steps in the correct order.
• + Captures the flat-vector infographic style perfectly with a cohesive NASA color palette.
• + Excellent text legibility and accuracy for main headers and crew names.

• − Minor spelling glitches in secondary text like '3rajcoory' and 'Transluiory'.
• − The icons for Earth Orbit and Lunar Orbit are very similar, both featuring rings.

OmniGen v2

• + Features a clean, modern layout with a strong grid system.
• + Adheres strictly to the requested muted red and navy color palette.

• − Fails to follow the sequential 6-step prompt, showing only 3 distinct iconography areas.
• − Severe spelling errors, including referring to the mission as 'Apolo 17' instead of Apollo 11.
• − The iconography is abstract and does not represent the requested Saturn V or Lunar Module effectively.