Grok Imagine Image Pro xAI vs Qwen Image 2.0 Alibaba

Grok Imagine Image Pro

• + Excellent photorealism with a highly believable wood texture and soft lighting.
• + Perfectly rendered glass physics, including accurate refraction of the plant pot behind it.
• + Includes legible and clever text on the book spine that relates to the prompt.

• − The plant pot is visible through the glass but the main plant body extends far above the cube.

Qwen Image 2.0

• + Follows all prompt elements including the glass cube, blue sphere, and red book.
• + Good use of soft window light from the left side.

• − The blue sphere appears to be floating unnaturally inside the cube rather than resting on the surface.
• − The refraction and reflections in the glass are messy, creating duplicate floating spheres that don't align with reality.
• − The composition feels a bit cramped with the cube very close to the window edge.

Grok Imagine Image Pro

• + Excellent full-body composition and spatial grounding
• + Captures the motion blur of passing cars clearly as requested
• + Highly realistic textures on the road and pavement environment

• − The hands and wrench interaction is slightly anatomically awkward
• − The lighting on the man's face is a bit flat compared to the backgrounds

Qwen Image 2.0

• + Superb skin texture and facial realism
• + Accurately represents the 'imperfect framing' and '50mm' feel with a tighter crop
• + Stronger puddles and rain-on-pavement details

• − The man's right hand is mangled with extra fingers and fused forms
• − The bicycle geometry is slightly warped near the pedal and chain area

Grok Imagine Image Pro

• + Excellent execution of the engraved plate armor and readable Latin text.
• + Superior detail in facial features, skin texture, and lifelike eyes.
• + Masterful use of lighting and depth of field with realistic bokeh sparks.

• − The hair is somewhat styled in a way that looks more like modern dreadlocks than a classic braid.

Qwen Image 2.0

• + Successfully captures a more rugged, older character that feels 'battle-worn'.
• + Effective use of beads in the hair braids and vibrant secondary colors.

• − The hand modeling is poor, with distorted fingers and anatomy.
• − The armor engraving is significantly less detailed and more generic than Model A.
• − The image quality is grainier with more visible artifacts.

Grok Imagine Image Pro

• + Excellent adherence to the requested grid and section layout.
• + Features readable and semi-coherent English text for item names and descriptions.
• + High-quality food photography with consistent lighting and vibrant colors.

• − The placeholder text for several pizza items is repeated exactly.
• − Some minor text artifacts in the smaller descriptions.

Qwen Image 2.0

• + Features a trendy, clean aesthetic with rounded corners on images.
• + Includes price indicators which enhance the menu feel.

• − Complete failure to organize by sections (labeled 'Mains' contains a pizza).
• − The text is entirely illegible garble.
• − Poor grid consistency with some images spanning columns and others not.

Grok Imagine Image Pro

• + Excellent dynamic motion with high-quality splashes and floating components.
• + Perfect text rendering for all three requested elements.
• + High level of photographic texture on the meat patty and melting cheese.

• − The starburst element looks a bit like a flat clip-art sticker compared to the rest of the image.
• − Price uses a comma instead of a period, though common in some regions.

Qwen Image 2.0

• + Text has a very impressive 'fiery' glow effect that integrates into the scene.
• + Effective use of vertical space and flame effects.
• + Clean and clear price starburst.

• − The burger is not clearly 'exploded' as much as Model A; the bottom half remains mostly stacked.
• − The text 'LIMITED TIME ONLY' is somewhat small and less prominent.

Grok Imagine Image Pro

• + Excellent text rendering with perfect spelling and placement.
• + Realistic chalk texture and natural variations in handwriting style.
• + High contrast and clear visibility of all requested menu items.

• − The composition is a bit flat and strictly centered, lacking a bit of environmental depth.

Qwen Image 2.0

• + Beautiful environmental lighting and composition that conveys the 'cozy café' atmosphere well.
• + Good use of chalk smudges and texture to enhance the realism of the board.

• − The text layout is slightly messy, particularly around the prices.
• − Multiple instances of small text artifacts and slightly less consistent handwriting compared to Model A.

Grok Imagine Image Pro

• + Perfect adherence to the unusual 'horse on top' request
• + Vibrant, cinematic colors and lighting
• + High detail in the nebula and planetary background

• − The horse appears to be floating just above rather than 'riding' in a traditional physical sense

Qwen Image 2.0

• + High textural detail on the horse and space suit
• + Good composition with Earth in the background

• − Completely failed the negative constraint/specific instruction of 'horse on top'
• − The scales on the horse's neck look a bit muddy and inconsistent

Grok Imagine Image Pro

• + Excellent composition from the dashboard perspective that clearly places the human in the back seat.
• + Very high level of photorealism and texture, particularly on the capybara's fur and the woman's clothing.
• + Accurate text rendering on the hat with 'NYC TLC' identifying the taxi context.

• − The capybara's hands look slightly more like paws with human-like finger articulation which is a bit uncanny.
• − The lighting is slightly flat compared to the cinematic night feel of model B.

Qwen Image 2.0

• + Dynamic angle that gives a great sense of being inside the vehicle with the driver.
• + The blurred city lights reflection on the window adds a high degree of realism to the night setting.
• + Great adherence to the 'bored expression' prompt for the passenger.

• − The passenger is physically sitting in the front passenger seat next to the driver, failing the prompt's request for her to be in the back seat.
• − The capybara's hat is generic and lacks the specific 'taxi driver' branding requested.
• − Lower overall resolution and clarity compared to Model A.

Grok Imagine Image Pro

• + Perfect text rendering for all requested details including date, time, and location.
• + Dynamic and moody cinematic lighting with smoke effects and varying color tones.
• + Excellent composition with a detailed thorn and web border integrated into the parchment design.

• − The text at the bottom is slightly crowded compared to the top title.

Qwen Image 2.0

• + Clear and legible gothic typography for the header.
• + A clean, symmetrical layout that works well for a square invitation.
• + Good adherence to the request for twisted trees and a thorn border.

• − The jack-o-lantern lighting is a bit flat compared to the other model.
• − The text in the small banner is slightly distorted and contains a comma where none was requested.
• − The overall image lacks the rich, atmospheric depth of the first image.

Grok Imagine Image Pro

• + Excellent 3D miniature 'cartoon' aesthetic as requested.
• + Precise 45-degree isometric perspective.
• + Very clean typography and creative flag icon integration.

• − Texture of the nigiri rice looks slightly uniform and synthetic.

Qwen Image 2.0

• + High realism in food textures, specifically the eel and salmon.
• + Clear, bold text rendering.
• + Good adherence to the square format and diorama base.

• − Missed the 'cartoon' style request, leaning towards a photographic look.
• − The flag icon is a standard rectangle rather than a stylized icon.
• − Perspective is a bit flatter than a true 45-degree isometric view.

Grok Imagine Image Pro

• + Excellent depiction of god rays and sunrise lighting consistent with the prompt.
• + Sharp focus on all subjects with very clean 'masterpiece' aesthetic.
• + Included all requested animal types clearly.

• − Generated two kittens instead of the requested one tabby kitten.
• − The butterflies look somewhat static and pasted-on compared to the environment.

Qwen Image 2.0

• + Correctly included exactly one of each animal requested.
• + The 'tumbling' interaction is much more natural and dynamic than in Image A.
• + The fur textures appear softer and more integrated with the lighting.

• − The fox's face/eye area shows slight anatomical warping during the tumble.
• − The bunny looks a bit detached from the main action in the center.

Grok Imagine Image Pro

• + Excellent typography and precise text rendering
• + Clean vector emblem style that looks like a professional logo
• + Well-balanced composition with an appropriate 'Est. 1720' banner

• − Steam effect is a bit simple/swirly rather than natural

Qwen Image 2.0

• + Stronger vector texture and gradients on the cloche dome
• + Creative integration of steam inside the dome's reflection area
• + Maintains warm brown and cream tones as requested

• − The ribbon/banner tails are asymmetrical and messy
• − The text alignment inside the ribbon is slightly off-center
• − Typography is less elegant and 'classic' compared to Model A

Grok Imagine Image Pro

• + Excellent typography with no spelling errors.
• + Consistent and clean flat-vector iconography for every step.
• + Perfect adherence to the requested NASA-inspired color palette and layout.

• − Step 3 (Translunar) icon is a bit abstract compared to the others.

Qwen Image 2.0

• + Strong composition with a sense of scale for the landing phase.
• + Good adherence to the requested steps.

• − Includes a spelling error ('Translunjar').
• − Inconsistent icon styles, mixing silhouette humans with detailed line-art modules.
• − Poor text placement where orbit rings overlap the 'Lunar Orbit' text.