Grok Imagine Image xAI vs Vidu Q2 ShengShu Technology

Grok Imagine Image

• + Excellent photorealism and depth of field
• + Accurate representation of refraction through the glass cube
• + Subtle and natural lighting that matches the 'soft window light' prompt

• − The cube geometry is slightly elongated, appearing more like a rectangular prism
• − The blue sphere is levitating rather than sitting on the bottom of the cube

Vidu Q2

• + Perfect cube geometry with clear edges
• + Higher detail on the book spine and plant textures
• + Realistic physics with the ball resting on the glass surface

• − The lighting is a bit harsh and contrasted compared to the 'soft' light requested
• − The reflection of the ball on the bottom surface looks more like a mirror than simple glass

Grok Imagine Image

• + Excellent photographic quality and motion blur
• + Perfect adherence to the environment requested
• + High-quality rendering of the Rolls-Royce convertible

• − Failed to use the specific man provided in the source image
• − The driver is a generic older Caucasian man instead of the young Black man in the reference

Vidu Q2

• + Successfully preserved the identity of the specific man from the source image
• + Good preservation of the car's design and features
• + Dynamics of the shot feel appropriate for a coastal drive

• − The driver is positioned on the wrong side for US roads (right-hand drive)
• − Technical quality is slightly lower than Model A, with some harsh edges around the driver

Grok Imagine Image

• + Excellent depiction of motion blur with a passing car as requested.
• + Highly convincing 'imperfect framing' that mimics a real candid photograph.
• + Strong adherence to the 50mm lens and shallow depth of field instructions.

• − The man's face is obscured, making it hard to judge age/ethnicity details.
• − The presence of a face mask feels slightly dated/specific, though realistic for Japan.

Vidu Q2

• + Exceptional skin texture and detail on the man's hands and face.
• + The bicycle mechanics look more varied and detailed.
• + Good representation of wet pavement and light rain.

• − Lacks the requested motion blur for the passing car.
• − The composition feels more staged/professional than the 'candid street photo' requested.
• − Some anatomical issues with the proximity and angle of the hands relative to the body.

Grok Imagine Image

• + Exquisite engraving detail on the plate armor
• + Beautiful atmospheric lighting and bokeh sparks that frame the subject
• + Strong adherence to the 'beads in hair' prompt with multiple visible beads

• − The armor looks almost too clean and polished despite the facial scars
• − Hair texture looks a bit soft and blurry in some areas

Vidu Q2

• + Superior texture on the cloth underlayer and leather straps
• + The armor reflects a much better 'battle-worn' aesthetic with grime and scratches
• + Very sharp and lifelike skin textures and facial anatomy

• − The 'braided with small beads' requirement is less prominent
• − The torchlight reflection on the metal is a bit more muted than Image A

Grok Imagine Image

• + Strictly follows the requested sections for Appetizers, Pizza, and Mains.
• + Excellent layout with a clear hierarchy and readable bold sans-serif fonts.
• + Integrated high-quality food photography that fits a professional menu style.

• − Contains repetitive text entries like multiple instances of 'Steak Frites' and 'Grilled Salmon'.
• − Minor spelling errors in larger text such as 'Calmon' and 'Saldes'.

Vidu Q2

• + Includes price placeholders which adds to the authenticity of a menu layout.
• + Vibrant and high-resolution food photography with good color saturation.

• − Layout is cluttered and lacks a clear grid structure for images.
• − The text is largely unreadable with heavy garbling and nonsensical words like 'APECIZEN' and 'PPEIZEIS'.
• − Failed to follow the specific sections requested, merging appetziers and pizza haphazardly.

Grok Imagine Image

• + Excellent typography with clean, flaming textures
• + Dynamic debris and sauce splashes add a great sense of motion
• + Perfect price tag rendering with the requested currency symbol

• − The starburst graphic for the price is slightly more generic than the rest of the art

Vidu Q2

• + Strong fiery atmosphere and background integration
• + The burger texture looks very appetising and detailed

• − The pricing text contains a symbol error, showing a pound/double-cross hybrid instead of a Euro symbol
• − The font choice for 'Limited Time Only' is less impactful than the headline text

Grok Imagine Image

• + Excellent text rendering with perfect spelling and legibility
• + Authentic chalk texture and smudging on the chalkboard
• + Effective completion of the truncated prompt description

• − The 'elegant cursive' for the title is more of a print-script hybrid than true cursive

Vidu Q2

• + Very realistic chalk strokes with layered pressure and varying thickness
• + Dynamic, hand-drawn composition that feels truly organic

• − Numerous spelling errors including 'Truffe Musshoom' and 'Octopd'
• − The text becomes garbled and illegible toward the bottom of the board
• − Incorrect price rendering for the first item

Grok Imagine Image

• + Slightly modifies the leg position to be more dynamic and balanced.
• + Maintains the high color saturation of the original environment.

• − Completely failed the character transfer, keeping the person from the pose reference.
• − Applied very minimal changes to the source image, largely ignoring the second input image.

Vidu Q2

• + Excellent character transfer, including sunglasses, scarf, and specific clothing details.
• + Perfectly replicates the complex leg-crossing pose and overall body position from Image 1.
• + Accurately integrates the character into the lighting and background of the target scene.

• − Slight distortion in the anatomy of the hands and fingers.
• − The watch on the wrist is slightly blurry compared to other details.

Grok Imagine Image

• + Excellent adherence to the 'horse on top' spatial instruction
• + High cinematic quality with beautiful nebula lighting
• + Strong surreal atmosphere

• − The horse's front hoof is oddly merged with the astronaut's hand

Vidu Q2

• + Vibrant colors and creative galaxy skin on the horse
• + Sharp focus and high detail resolution

• − Completely failed the semantic instruction of 'horse on top'
• − Generic composition that ignores the 'surreal' inversion requested

Grok Imagine Image

• + Excellent preservation of the person's identity and facial features
• + High quality rendering of the new garment

• − Completely ignored the clothing in Image 2, choosing a random regal outfit instead

Vidu Q2

• + Successfully transferred the correct clothing/accessories from Image 2
• + Maintains the background and general aesthetic of Image 1

• − Significant distortion of the model's facial shape and features
• − Inconsistent arm rendering with sand textures appearing as part of the skin or jacket
• − Modified the hairstyle and added a mustache contrary to instructions

Grok Imagine Image

• + Excellent adherence to the 'bored' expression for the passenger
• + High-quality fur texture and realistic lighting on the capybara
• + Captures the New York night atmosphere well with recognizable street elements

• − The passenger is sitting in the front passenger seat instead of the back seat
• − The capybara's hands look more like paws with prominent claws which might be slightly unsettling

Vidu Q2

• + Correctly places the human passenger in the back seat as requested
• + Features a more detailed driver cap with a badge
• + The capybara's paws look surprisingly natural on the steering wheel

• − The passenger's expression is slightly more vacant than 'bored'
• − Slightly less realistic skin texture on the human compared to Model A

Grok Imagine Image

• + Perfect text rendering for all requested titles and event details.
• + Highly cohesive composition with a well-balanced border of thorns and webs.
• + Superior atmosphere with cinematic lighting and a moody gothic background.

• − The transition between the dark background and the parchment edges is slightly abrupt.

Vidu Q2

• + Dynamic border elements and a nice illustration style for the Jack-o-lantern.
• + Captures the scroll banner and gothic aesthetic reasonably well.

• − Significant text errors including misspellings in 'Invitation' and the banner.
• − Incorrect date and time details, failing the specific accuracy of the prompt.
• − Lacks the atmospheric depth and background details like the moody sky and twisted trees requested.

Grok Imagine Image

• + Excellent source preservation across the entire image.
• + Realistic hair texture and color that matches the existing beard.
• + Natural, believable hairline that integrates well with the forehead.

• − The hair volume is slightly conservative compared to the request for a 'full, thick' head of hair.

Vidu Q2

• + Successfully added a very thick and voluminous head of hair as requested.
• + Matches the hair color and lighting of the scene quite well.

• − The hairline appears slightly artificial and too straight across the forehead.
• − Minor distortion to the forehead shape compared to the source image.

Grok Imagine Image

• + Excellent typography with clean, centered alignment
• + Sophisticated 3D rendering with realistic PBR textures on the fish and wood
• + Perfectly adheres to the isometric 45° perspective

• − The flag icon is a bit large compared to the text

Vidu Q2

• + Creative use of the flag being held by the text
• + Good colors and soft lighting

• − The text is not perfectly centered and looks slightly wonky
• − Sushi components have less defined textures compared to Model A
• − Perspective is slightly more frontal than the requested 45° isometric

Grok Imagine Image

• + Excellent caricature style with a classic large-head-small-body exaggeration.
• + Strong storytelling by placing the character at a news desk with thematic elements on screen.
• + High textual clarity and professional lighting effects.

• − The hands on the desk are very poorly rendered and look deformed.
• − The eyes feel slightly more robotic than the original subject's expression.

Vidu Q2

• + Maintains the subject's original clothing (denim shirt and black top) more accurately.
• + Highly creative integration of dogs, including a small dog in a hockey jersey.
• + Better hand rendering, despite the stylized four-finger cartoon approach.

• − The facial likeness is slightly weaker and more generic than Model A.
• − The microphone is quite massive and overlaps the hair, creating a cluttered composition.

Grok Imagine Image

• + Strong dynamic composition that emphasizes the 'tumbling together' action
• + Excellent lighting effects with clear God rays and golden hour ambiance
• + High level of softness and 'fluffiness' in the textures

• − Failed to include the butterflies mentioned in the prompt
• − Art style leans toward 3D digital illustration rather than 'hyper-photorealistic'
• − Anatomy is slightly stylized with oversized eyes

Vidu Q2

• + Accurately included all requested animals plus the butterflies
• + Achieved a much more 'hyper-photorealistic' look than the competitor
• + Spacious composition captures the 'lush wildflower meadow' perfectly

• − Lighting is a bit flat compared to the dramatic rays in Model A
• − The second golden retriever puppy was not requested (prompt asked for one)

Grok Imagine Image

• + Successfully captures the hand-painted, watercolor texture characteristic of Studio Ghibli backgrounds.
• + Maintains high character fidelity to the original meme while translating features into the requested art style.
• + Creates a cohesive, nostalgic mood with soft lighting and puffy clouds.

• − The facial expression of the man is slightly more subdued than the original meme.

Vidu Q2

• + Excellent use of pastel colors and dreamy, bright lighting that fits the 'aesthetic' side of anime art.
• + Preserves the iconic facial expressions of the 'distracted boyfriend' meme very effectively.
• + Includes nice line-art details on the clothing and hair.

• − The background feels a bit more like generic digital watercolor rather than the specific, lush Ghibli painting style.
• − The woman in the red dress has slightly inconsistent facial lighting compared to the rest of the scene.

Grok Imagine Image

• + Excellent wind effect on the hair that feels natural and dynamic.
• + Significant volume of leaves adds a strong sense of chaotic motion.
• + Highly accurate preservation of the subject's face and the dog.

• − The leaves appear somewhat repetitive in shape and color.
• − Almost all leaves were changed to orange, losing the original green summer feel of the park.

Vidu Q2

• + Natural integration of leaves with varied motion blur and sizes.
• + Beautifully rendered hair movement that maintains flow and realism.
• + Excellent preservation of the original image identity and colors.

• − Fewer leaves across the scene compared to Model A, resulting in slightly less 'energetic' feel.
• − Minor distortion on a few leaves in the foreground.

Grok Imagine Image

• + Perfect text rendering of the restaurant name and establishment date
• + Clean vector illustration with professional logo composition
• + Effective use of the requested brown and cream color palette with a subtle grain texture

• − The addition of a spoon and cup handle merging with the cloche makes the silhouette slightly cluttered

Vidu Q2

• + The banner design is classic and aesthetically pleasing
• + Follows the light background and warm tone instructions well

• − Significant spelling errors in the brand name and 'Est' text
• − The steam effect inside the cloche appears somewhat muddy and lacks clarity
• − Messy typography with inconsistent letter sizing and odd characters

Grok Imagine Image

• + Excellent text legibility for the main headings and crew names.
• + Closely adheres to the specified NASA-inspired color palette and flat-vector style.
• + Contains all six requested steps in a logical, numbered sequence.

• − The 'Translunar' section has visual artifacts and garbled sub-text.
• − The Earth icon mistakenly includes a ring, which is traditionally associated with Saturn.
• − The NASA logo is a slightly deformed approximation rather than a clean vector.

Vidu Q2

• + Features a clean, high-resolution aesthetic with consistent iconography.
• + Good use of the light gray and navy colors from the requested palette.
• + Stronger interpretation of the translunar trajectory and lunar surface 3D perspective.

• − Text is largely nonsensical or misspelled (e.g., 'ALFONCH', 'Aeutth Orot').
• − Fails to follow the requested numbering for the six specific steps.
• − The icons for descent and landing are repetitive rather than distinct steps.