LongCat-Image Meituan vs Vidu Q2 ShengShu Technology

LongCat-Image

• + Excellent depiction of soft window lighting from the left
• + Clean and realistic glass refractions
• + Good spatial arrangement of the plant behind the object

• − The sphere appears slightly large relative to the prompt 'small'

Vidu Q2

• + Complex and realistic shadows on the wooden table
• + High detail on the book spine and plant leaves
• + Accurate sphere size as requested

• − The lighting is very harsh and direct, contradicting the request for 'soft' light
• − The cube edges have some visual artifacts where they meet the book

LongCat-Image

• + Excellent wide composition that tells a story.
• + Strong adherence to the rain, reflections, and car lights prompt.
• + The red bicycle is a central, vibrant focal point.

• − Internal logic issues with the bicycle anatomy, showing extra wheels and strange frame connections.
• − Rain effects look slightly like static lines rather than natural droplets.

Vidu Q2

• + Exceptional skin and hand texture, appearing very realistic.
• + Stronger 'candid' feel with an tight, imperfect crop.
• + Excellent mechanical rendering of the chain and gears.

• − Misses the 'motion blur from passing cars' request as the car is static and sharp.
• − The overall image feels a bit crowded compared to the requested 50mm cinematic look.

LongCat-Image

• + Excellent execution of bokeh sparks and warm firelight reflection
• + High-quality rendering of the engraved patterns on the breastplate
• + Clearly visible beads in multiple braids as requested

• − The facial wounds appear somewhat artificial, like paint rather than deep scars
• − The armor lacks the 'battle-worn' texture, appearing very polished

Vidu Q2

• + Superb texture work on the cloth underlayer and leather straps
• + The facial scars and skin texture look more realistic and integrated
• + Effective use of warm torchlight to define the armor's shape

• − The braiding is less prominent and does not feature 'small beads' effectively as requested
• − The sparks in the background are less distinct compared to the other model

LongCat-Image

• + Strong use of vibrant accent colors that divide the page well
• + Clear grid-based layout that matches the prompt's structural request

• − The font is overly stylized and difficult to read
• − Includes a very large, unnecessary logo/header area that wastes space

Vidu Q2

• + Features a cleaner, more professional sans-serif typeface
• + Better organization of price points and item descriptions which feels more like a real menu
• + Higher clarity in the food photography shown in the grid

• − The 'Apecizen' and 'Appeczyers' headings are repetitive and misspelled
• − Slightly less bold use of accents compared to Model A

LongCat-Image

• + Excellent text rendering with no spelling errors
• + High photographic clarity and clean lighting
• + Professional ad layout with a distinct starburst element

• − The burger is not exploded as requested; it is a solid stack
• − Lacks the sense of motion for the individual ingredients

Vidu Q2

• + Successfully follows the ‘exploded’ component instruction
• + Dynamic background with great motion and fiery energy
• + More creative interpretation of the suspended ingredients

• − Failed to render the Euro symbol correctly, using a strange hybrid character
• − Text integration is a bit cluttered at the top

LongCat-Image

• + Excellent chalk-like texture on the board surface
• + Atmospheric background and realistic lighting
• + Compositionally well-balanced frame

• − Abysmal text rendering with numerous spelling errors and gibberish
• − The date is missing the 'P' in April
• − Failed to include the specific third menu item requested

Vidu Q2

• + Much better text legibility and spelling overall
• + Captured the third menu item (Brown Butter) mentioned in the prompt
• + Excellent handwritten cursive style with authentic chalk smudges

• − Several spelling errors like 'Musshoom', 'Risoto', and 'Lemepun'
• − The price for the first item was changed to $34 instead of the requested $24
• − Price for the third item became distorted symbols

LongCat-Image

• + Features a distinct landscape that gives a sense of scale and cinematic lighting.
• + The horse's anatomy and texture provide a grounded, high-detail feel.

• − Failed the specific instructional prompt 'horse on top, not vice versa'.
• − Includes distracting artifacts like a distorted airplane in the starfield.

Vidu Q2

• + Vibrant color palette with ethereal, cosmic lighting.
• + Artistically interpreted 'space' by making the horse itself part of the nebula.

• − Failed the specific structural instruction 'horse on top, not vice versa'.
• − The harness/reins are messy and lack physical coherence.

LongCat-Image

• + Features a high-quality capybara head with realistic fur texture.
• + The lighting on the taxi exterior and roof sign is vibrant and realistic.

• − The capybara's hand is depicted with human-like fingers and long claws, creating a creepy anatomical hybrid.
• − Only one paw is on the steering wheel, failing the prompt requirement.
• − Includes two passengers instead of the requested single businesswoman.

Vidu Q2

• + Perfectly adheres to the prompt by having both paws on the steering wheel.
• + The composition accurately captures the 'inside the taxi' perspective requested.
• + The businesswoman's bored expression and single passenger status correctly match the prompt.

• − The capybara's face is slightly less detailed and looks somewhat composited compared to the background.
• − Minor perspective warping on the interior dashboard.

LongCat-Image

• + Excellent text rendering with almost perfect spelling for the main title and banner.
• + Strong composition with a creative cutout effect in the parchment to show the landscape.
• + Sharp visual quality and high-contrast cinematic lighting.

• − Misspelled 'The Arches' as 'The Armiees' at the bottom.
• − The thorns are a bit repetitive and look like a digital overlay rather than integrated art.

Vidu Q2

• + Atmospheric integration of the twisted trees and spider webs into the border.
• + The parchment texture feels more authentic and vintage.

• − Significant spelling errors throughout, including 'Intovztion' and 'might of fiigts'.
• − Incorrect date and time formatting compared to the prompt (30.70.2025).
• − The central jack-o-lantern lacks the professional polish found in the other model.

LongCat-Image

• + Excellent typography with clean, bold execution.
• + Superior clay-like 3D textures that fit the 'miniature' prompt perfectly.
• + High-clarity lighting and shadows create a professional diorama feel.

• − The flag icon is stylized enough to appear slightly warped in its waving effect.

Vidu Q2

• + Includes a wider variety of sushi items.
• + Good adherence to the 45-degree isometric angle.
• + Features a playful flag icon integration.

• − The 'SUSHI' text is slightly misaligned with the 'JAPAN' text.
• − The surface details on the sushi look a bit more plastic and less 'refined' than the competitor.
• − The base contains some odd, indistinct small artifacts at the corners.

LongCat-Image

• + Strong implementation of god rays and golden hour lighting
• + Clear, high-quality textures on the fur of the main subjects
• + Vibrant colors and a very high-quality artistic finish

• − Serious anatomical failure, merging the kitten and bunny into a single 'cat-rabbit' hybrid
• − The butterfly and water droplets appear pasted on rather than integrated into the scene

Vidu Q2

• + Successfully includes all four distinct animal species requested
• + More dynamic composition showing the animals actually playing and leaping
• + Better integration of the butterflies and wildflowers within the environment

• − One of the puppies has a slightly distorted paw
• − Lower overall sharpness compared to the other model

LongCat-Image

• + Excellent typography rendering with almost perfect spelling
• + Strong woodcut/etched texture that fits the vintage theme
• + Well-integrated composition of the cloche and banner elements

• − Includes a repetitive 'Caffé' word within the graphic
• − The steam effect is a bit heavy-handed for a minimalist prompt

Vidu Q2

• + Elegant and clean vector-style cloche illustration
• + Sophisticated warm color palette and soft lighting
• + Good use of white space

• − Severe spelling errors in both the main title and 'Est.' banner
• − Includes redundant and malformed text at the bottom of the frame
• − Fails to correctly render the 'Florian' brand name

LongCat-Image

• + Stronger adherence to the navy and white NASA-inspired color palette
• + The landing illustration is detailed and thematic
• + Text rendering is bold and stylized, though largely illegible

• − Fails to include all 6 requested steps, showing only a few disconnected icons
• − Layout is cluttered and lacks clear logical flow for an infographic
• − The rocket icon looks more like a modern space shuttle than a Saturn V

Vidu Q2

• + Better adherence to the sequential 6-step structure requested in the prompt
• + Clean, modern flat-vector aesthetic with consistent iconography
• + Includes distinct icons for launch, orbit, translunar, and descent phases

• − Background is a bit pale/washed out compared to the 'navy' request
• − The text labels are gibberish despite attempting a numbered sequence
• − Icon for Earth Orbit and Lunar Orbit are very similar, lacking variety