Vidu Q2 ShengShu Technology vs Wan 2.5 (Preview) Alibaba

Vidu Q2

• + Perfect adherence to spatial prompt instructions, including the blue sphere inside and the plant behind.
• + Excellent rendering of glass physics, reflections, and wood grain texture.
• + Highly realistic lighting and shadows that align with the window source described.

• − The plant foliage is somewhat repetitive and looks slightly synthetic compared to the other objects.

Wan 2.5 (Preview)

• + High visual appeal with a cinematic depth of field and soft lighting.
• + Good texture on the red book, showing realistic wear on the edges.

• − Failed to place the sphere 'inside' the cube properly; it appears to be intersecting the glass or hovering in a broken space.
• − Distracting artifacts like floating dust motes that weren't requested.
• − The plant's position through the glass is distorted in a way that breaks the cube's geometry.

Vidu Q2

• + Excellent preservation of the car's specific make and model details.
• + Succesfully places the man from the second source image into the driver's seat.
• + Dynamic motion blur on the road enhances the 'driving' theme.

• − The man's hair appears slightly altered and simplified compared to the source.
• − The perspective of the road behind the car is slightly disjointed.

Wan 2.5 (Preview)

• + Very realistic lighting and integration of the car into the coastal environment.
• + Captures the iconic California coastline aesthetic with palm trees and cliffs.
• + Good motion blur on the wheels and road.

• − The man is very small and lacks detail, making him difficult to identify as the subject from the source.
• − Significant artifacts are visible on the man's face and hair.

Vidu Q2

• + Excellent skin texture and weathered detail on hands and face
• + Effective 'imperfect framing' that feels like a genuine candid snap
• + Great wet pavement reflections and realistic rain effects

• − The bicycle anatomy is physically impossible with multiple frames and mangled chains
• − Motion blur on the car is subtle/lacking compared to the prompt
• − Distracting artifacts near the man's forehead

Wan 2.5 (Preview)

• + Superb composition with deep reflections and visible raindrops
• + High clarity and realistic depth of field
• + Accurate depiction of a red bicycle and repair tools

• − Skin texture is slightly too smooth and looks a bit digital
• − Lacks the requested 'motion blur from passing cars'
• − The framing is very centered and professional rather than 'imperfect'

Vidu Q2

• + Excellent crispness in the ornate engravings on the plate armor.
• + Very detailed texture on the leather straps and the thick woven cloth underlayer.
• + Natural looking face with subtle scarring and consistent lighting.

• − The 'braided hair with beads' is a bit understated compared to the request.
• − Lighting feels slightly more studio-like rather than originating from a distinct torch.

Wan 2.5 (Preview)

• + Superior interpretation of the 'hair braided with small beads' and 'dirt on skin' prompts.
• + The warm torchlight and bokeh sparks are more prominent and atmosphere-enhancing.
• + Excellent realism in the eyes, showing redness and fatigue consistent with a 'battle-worn' character.

• − The plate armor engravings are slightly less sharp than Model A.
• − The tattered cloth edges look a bit more artificial compared to the woven texture in Model A.

Vidu Q2

• + Includes pricing column which adds to the menu realism
• + Vibrant and high-quality food photography
• + Good use of accent colors to separate sections

• − Text is largely nonsensical and features poor character rendering
• − Layout feels a bit cluttered and lacks the 'minimalist' feel requested

Wan 2.5 (Preview)

• + Closer adherence to the 'minimalist' aesthetic with clean spacing
• + Clearer text rendering for main headings
• + Excellent grid layout that separates the three requested sections

• − Food photos are repetitive and look very similar across different dishes
• − Included 'Modern minimalist' text at the top as if it were a title rather than a design style

Vidu Q2

• + Excellent typography style with a convincing fiery glow effect.
• + Bright, high-contrast colors that pop against the background.
• + Follows the general composition requirements for a food advertisement.

• − The currency symbol is incorrect, showing a crossed 'E' rather than a Euro sign.
• − The burger is not very 'exploded' as requested, with most components still touching.

Wan 2.5 (Preview)

• + Perfect execution of the 'exploded' burger request with dynamic separation of all ingredients.
• + Accurate text rendering including the correct Euro symbol (€).
• + Superior photorealistic detail on individual ingredients like the lettuce and tomato slices.

• − The 'MAGIC BURGER' text looks slightly drippy/melted rather than strictly fiery.
• − Layout is a bit more scattered, though this fits the 'exploded' prompt better.

Vidu Q2

• + Excellent realistic chalk texture with dusty smudges
• + Dynamic and natural-looking handwriting style

• − Numerous spelling errors including 'Lemepun' and 'Browd Botter'
• − Failed price accuracy on multiple items
• − Text becomes nonsensical at the bottom

Wan 2.5 (Preview)

• + Perfect adherence to text spelling and prices
• + Very clean legibility while maintaining a chalkboard aesthetic
• + Handled the complex brown butter chocolate chip cookie item well

• − The handwriting looks slightly too uniform, bordering on a digital font look
• − Missing the requested 'elegant cursive' for the title

Vidu Q2

• + Excellent visual coherence with a galactic texture on the horse.
• + Very high cinematic quality and vibrant colors.

• − Failed the negative constraint; the astronaut is riding the horse, not vice versa.
• − The horse's left front leg has an unnatural attachment to the body.

Wan 2.5 (Preview)

• + Clean details on the astronaut's suit and the horse's anatomy.
• + Good depth of field and composition with the planets in the background.

• − Failed the negative constraint; the astronaut is riding the horse.
• − The horse's tail and hind legs show motion blur that feels slightly disconnected from the foreground.

Vidu Q2

• + Successfully preserved the unique vitiligo patterns on the person's face and arm.
• + Accurately included the specific accessories like the watch and sunglasses from Image 2.
• + Maintained the body pose and beach environment of the source image.

• − The right sleeve is missing, exposing the arm which contradicts 'long-sleeve' layers from the outfit.
• − Added a mustache and changed the facial structure slightly.
• − Introduced sand artifacts on the coat that weren't in the original.

Wan 2.5 (Preview)

• + Successfully applied the full outfit including the coat, scarf, and watch.
• + Maintained the general background and lighting of Image 1.
• + High resolution and clear textures on the clothing.

• − Completely failed to preserve the base person's identity, replacing the model from Image 1 with the model from Image 2.
• − Lost the unique vitiligo features that were central to the 'base person' instruction.
• − Changed the person's hair and facial features entirely.

Vidu Q2

• + Excellent photorealistic lighting and textures
• + Perfectly captures the passenger's bored expression
• + The capybara's anatomy and fur are rendered with high fidelity

• − The driver's front paw has a slightly strange, hand-like structure
• − The view through the window is slightly less recognizable as Manhattan specifically

Wan 2.5 (Preview)

• + Strong composition with the taxi's exterior and interior visible
• + Excellent city background that feels uniquely like Times Square/NYC
• + Effective use of rain on the windshield for mood

• − The capybara's paws look more like strange human hands with long fingernails
• − The passenger's face is slightly blurry and less detailed than model A
• − The text on the taxi sign is nonsensical

Vidu Q2

• + Features a classic vintage parchment aesthetic
• + The border of thorns and webs is well-integrated with the artwork

• − Has multiple significant spelling errors in all text fields
• − Failed to include the correct date (30.70.2025 instead of 30.10.2026)

Wan 2.5 (Preview)

• + Excellent text rendering with no spelling mistakes in any of the fields
• + Superior visual depth with cinematic lighting and a clear moody sky inside a portal
• + Accurately followed all specific details including date, time, and location

• − The thorns on the border look a bit repetitive or pattern-like

Vidu Q2

• + Excellent texture and realistic hair strands
• + Preserves the original facial identity perfectly
• + Maintains consistent lighting across the new hair and the face

• − The hairline on the forehead is slightly too straight and abrupt

Wan 2.5 (Preview)

• + Natural, softer hair volume and styling
• + Excellent integration of temples and sideburns into the existing beard
• + Preserves details and background perfectly

• − The face undergoes a subtle shift in features, making him look like a slightly different person
• − The hair texture is a bit softer/blurry compared to the rest of the image

Vidu Q2

• + Excellent variety of sushi types on the plate
• + Accurate rendering of all requested text and the flag icon
• + High level of detail in textures and modeling

• − The diorama base has some strange, ambiguous small shapes on the surface
• − The flag is floating or attached poorly to the text

Wan 2.5 (Preview)

• + Perfectly clean, minimalist aesthetic that fits the 'soft refined textures' prompt
• + Very professional typography and flag placement
• + Superb lighting and soft-focus composition

• − Only depicts a single piece of sushi, which feels slightly empty for a 'scene'
• − The rice texture is slightly repetitive and round like beans

Vidu Q2

• + Excellent facial resemblance to the subject in the source image.
• + Creative integration of the hockey rink as the background and paw prints on the desk papers.
• + Preserves the specific denim shirt and black top from the original photo.

• − The fingers on the left hand are anatomically incorrect and poorly rendered.
• − The 'NEWS' tag on the microphone looks slightly amateurish in design compared to the rest of the illustration.

Wan 2.5 (Preview)

• + Stronger caricature style with classic 'big head' exaggeration.
• + Clearer inclusion of hockey elements, including a stick and a TV monitor showing a game.
• + Very clean and balanced composition suitable for a profile picture or avatar.

• − The facial resemblance to the source image is significantly weaker than the other model.
• − The hands holding the microphone are very small and lack detail.
• − The background is more generic and less integrated than the rink design in the other model.

Vidu Q2

• + Excellent composition showing motion and multiple animals without crowding.
• + Higher detail in the wildflower meadow and background lighting.
• + More accurate butterfly anatomy and variety.

• − Includes two golden retriever puppies instead of one as requested.
• − The bunny has a slightly strange tail/hind-leg configuration.

Wan 2.5 (Preview)

• + Captures the 'big expressive eyes' and 'ultra-detailed soft fur' perfectly.
• + Follows the count of animals accurately (one of each).
• + Beautiful rendering of dew drops and god rays.

• − The fox's eyes appear slightly unnatural and glowing.
• − The butterfly on the right has disconnected wings.

Vidu Q2

• + Excellent preservation of the source image's layout and character poses.
• + Effective use of watercolor-style textures that mimic traditional Ghibli background art.
• + Accurately captures the specific facial expressions from the original meme.

• − The character line art is a bit inconsistent in thickness compared to the background.
• − Stronger reliance on digital-looking shading in some areas of the clothing.

Wan 2.5 (Preview)

• + Stronger Ghibli aesthetic with the addition of characteristic floating leaves and soft bloom.
• + The character faces are more successfully stylized into the specific Ghibli 'look'.
• + Beautiful soft lighting and color palette that matches the 'dreamy' prompt.

• − The woman on the right has a slightly more neutral expression, losing some of the 'distracted boyfriend' meme's humor.
• − Some minor loss of structural detail in the background architecture.

Vidu Q2

• + Excellent adherence to the 'leaves flying' instruction with realistic autumn colors.
• + Successfully animated the hair to look wind-blown while keeping facial features intact.
• + Excellent source preservation, keeping the person and dog backgrounds virtually unchanged.

• − The amount of leaves is slightly overwhelming, partially obscuring the background landscape.

Wan 2.5 (Preview)

• + Natural and elegant wind-blown hair effect that looks very lively.
• + Good source preservation of the human subject and the dog.
• + Subtle use of leaves preserves the original composition well.

• − The green leaves look like digital stickers/overlays rather than integrated objects.
• − The leaves lack the lighting and texture detail found in Model A.

Vidu Q2

• + Successfully captures the warm brown and cream color scheme.
• + The illustration of the cloche dome includes nice stylized steam.

• − Significant text errors including 'CAFFE FARMIIN' and 'Esttt'.
• − The inclusion of a secondary, garbled logo line below the main banner is redundant and messy.
• − Vector lines are a bit soft/blurry in places.

Wan 2.5 (Preview)

• + Excellent typography with perfect spelling of 'CAFFÈ FLORIAN'.
• + Clean, high-quality vector style that fits the minimalist emblem request.
• + Superior background texture that mimics vintage paper.

• − The cloche dome is a bit small relative to the banner size.
• − The steam inside the dome is a bit thick, losing some delicate detail.

Vidu Q2

• + Uses a clean, flat-vector style for the icons.
• + Follows the color palette requested with muted tones.

• − Text is largely gibberish and full of spelling errors.
• − The icons for the landing phase are repetitive and lack distinct 'descent' vs 'landing' variation.
• − The layout is somewhat cluttered and lacks a clear chronological flow.

Wan 2.5 (Preview)

• + Excellent text rendering with correct spelling for labels and astronaut names.
• + Logical infographic flow that traces the trajectory from Earth to Moon.
• + Strong adherence to the 'NASA-inspired' navy and white color palette.

• − The Saturn V rocket icon is used twice, even for the descent phase where a lunar module was requested.
• − One of the astronaut icons appears to be wearing a helmet that resembles a Stahlhelm rather than a flight suit.