GPT Image 1 OpenAI vs Vidu Q2 ShengShu Technology

GPT Image 1

• + Excellent photographic clarity and soft lighting
• + Clean minimal composition
• + Accurate representation of the blue sphere inside the cube

• − The cube geometry is slightly warped at the corner joints
• − Reflection of the sphere on the base is missing or inconsistent

Vidu Q2

• + Detailed textures on the book and table
• + Accurate implementation of shadows and reflections within the glass cube
• + Rich, dynamic lighting effect from the window

• − The plant leaves are slightly blurred and look less realistic than the foreground
• − The plant's pot is hidden in a way that makes it look like it is floating

GPT Image 1

• + Excellent character preservation of the man's face and hairstyle.
• + Natural-looking lighting and motion blur that blends the car into the environment.
• + Maintains the specific proportions and model details of the source car.

• − The driver side is on the right, whereas both the source car and American roads typically use a left-hand drive configuration.

Vidu Q2

• + Successfully places the driver on the left side of the vehicle.
• + Produces a vibrant and clear landscape for the California coastline.

• − Significant loss of character likeness; the man's hair and facial features are generic.
• − Noticeable artifacts around the front grill and headlights of the car.
• − The perspective of the road lines and the car's motion feels slightly distorted.

GPT Image 1

• + Excellent skin texture and realistic face weathering
• + Beautifully rendered rain droplets on the man's hair and clothing
• + Strong atmospheric perspective with soft bokeh in the background

• − The rear wheel of the bicycle lacks a chain and has nonsensical gear structures
• − The car in the background is static with no motion blur as requested

Vidu Q2

• + Successfully captures 'imperfect framing' with a close-up, candid feel
• + Shows complex bicycle mechanics like the chain and derailleur
• + Reflections on the wet pavement are vibrant and detailed

• − Anatomical issues with three hands appearing to work on the bike
• − The background car lacks the requested motion blur
• − The subjects are partially cropped out in an awkward way

GPT Image 1

• + Exceptional skin texture with realistic dirt and bloodstains
• + Highly ornate and weathered engraving on the plate armor
• + Atmospheric lighting that perfectly captures the warm torchlight and bokeh sparks

• − The beads in the hair are very subtle and blend in with the braids
• − The leather straps mentioned in the prompt are largely obscured

Vidu Q2

• + Excellent depiction of the leather straps and garment textures
• + Clearly visible beads and ornaments in the braided hair
• + Sharp focus on the face and scars

• − The skin and hair look slightly too clean for a battle-worn character
• − The lighting feels more like studio lighting with a warm filter rather than natural torchlight
• − Armor textures lack the fine-etched detail seen in the competitor

GPT Image 1

• + Excellent typography with legible bold sans-serif fonts
• + High-quality, realistic food photography
• + Clean minimalist layout that looks professional

• − Repetitive placeholder text with typos like 'descrigion'
• − Limited layout complexity compared to a full menu

Vidu Q2

• + More comprehensive menu layout with multiple columns
• + Good use of color accents throughout the design
• + Captures the 'vibrant' and 'casual dining' atmosphere well

• − Garbled, unreadable text throughout the entire image
• − Messy, incoherent food images with many digital artifacts
• − Chaotic layout that lacks the requested 'minimalist' feel

GPT Image 1

• + Excellent typography with a consistent glowing, fiery texture.
• + Photorealistic food textures, especially the sear on the patty and the freshness of the lettuce.
• + Clean composition with balanced space for all required text elements.

• − Incorrect price rendering, showing '€.99' instead of '€6.99'.
• − The starburst shape is a bit jagged and lacks the 'fiery' rendering of the other text.

Vidu Q2

• + Dynamic and exciting background with intense fire effects and motion.
• + Accurate numerical price rendering within a comic-style starburst.
• + Good integration of the 'LIMITED TIME ONLY' text with the main title.

• − Poor character rendering; the Euro symbol and several letters in 'LIMITED' are mangled.
• − The bottom bun texture appears blurry and low-resolution compared to the top half.
• − The lettuce and tomato layers look somewhat artificial and flat.

GPT Image 1

• + Excellent text legibility and spelling accuracy
• + Consistent chalk texture across all letters
• + Followed the specific menu items and prices almost perfectly

• − Handwriting feels slightly 'font-like' and uniform compared to real chalk art
• − Missed the 'elegant cursive' requirement for the title

Vidu Q2

• + Features a very authentic, artistic 'hand-drawn' chalk style with natural smudges
• + Captures the cozy café background environment
• + Good use of varying chalk weights for emphasis

• − Significant spelling errors throughout the menu items
• − The handwriting is messy to the point of being difficult to read
• − Failed to maintain the requested price for the first item

GPT Image 1

• + Successfully replicates the complex cross-legged pose on the red box
• + Captures the lighting and yellow background of source image 1 well

• − Face and features bear little resemblance to the character reference in Image 2
• − The hands and lower arm anatomy are distorted and lack detail
• − The scarf texture is flat and lacks the realistic draping of the original

Vidu Q2

• + Excellent character preservation with high facial similarity to Image 2
• + Accurately replicates the scarf pattern and clothing details from the character reference
• + Precise adherence to the dynamic pose/anatomy while maintaining higher image resolution

• − None significant; captures all aspects of the multi-image instruction effectively

GPT Image 1

• + Excellent cinematic lighting and realistic textures on the horse's coat and spacesuit.
• + Clean composition with a consistent mood.
• + High level of anatomical detail in the horse's musculature.

• − Failed the specific spatial instruction 'horse on top, not vice versa'.
• − Interpretation is conventional rather than truly surreal.

Vidu Q2

• + Vibrant, imaginative use of nebulae themes within the horse's skin.
• + Captured the 'surreal' aspect of the prompt more effectively through color and lighting.
• + Dynamic composition with a sense of motion.

• − Failed the specific spatial instruction 'horse on top, not vice versa'.
• − Some minor artifacts where the astronaut's hands interact with the reins.
• − Colors are somewhat oversaturated, leaning more toward digital art than cinematic realism.

GPT Image 1

• + Excellent preservation of the person's facial features and vitiligo patterns
• + High-quality rendering of the clothing and scarf textures
• + Maintains the lighting and atmospheric feel of the original beach scene

• − The scarf pattern, while similar, is not an exact match to the source
• − The watch added is a different style than the gold one in Image 2

Vidu Q2

• + Successfully includes the sunglasses and gold watch from Image 2
• + Adds the scarf and coat with reasonable accuracy

• − Significantly alters the person's face and hair, failing the 'completely unchanged' constraint
• − Major anatomical errors including a missing right arm and sand texture floating over clothing
• − Poor lighting integration making the figure look like a cutout

GPT Image 1

• + Excellent photorealistic texture on the capybara's fur
• + Cinematic lighting and composition that creates a moody atmosphere
• + Captures the bored, non-plussed expression of the passenger perfectly

• − The passenger appears to be in the front passenger seat rather than the back seat
• − The capybara's paws look more like primate hands than capybara paws

Vidu Q2

• + Accurately places the passenger in the back seat as requested
• + Bright, clear details of the taxi interior and city lights
• + The capybara's expression is very professional and alert

• − The capybara appears to be floating or disconnected from the headrest
• − The passenger's phone and fingers have some minor distortion
• − The lighting feels a bit more artificial compared to Model A

GPT Image 1

• + Perfect text rendering for all requested fields with no spelling errors.
• + Excellent composition that feels professional and balanced.
• + Atmospheric lighting and textures that evoke a true vintage gothic aesthetic.

• − The fonts, while elegant, are more serif-classic than distinctly 'gothic' compared to Model B.
• − The border is a bit subtle, blending into the dark background.

Vidu Q2

• + Captures the 'gothic' typography style beautifully for the main title.
• + Includes a clear parchment-style background as requested in the prompt.
• + Vibrant lighting on the Jack-o-lantern creates a strong focal point.

• − Extremely poor text accuracy with numerous spelling errors like 'Intovztion' and 'might of fiigts'.
• − Incorrect date (2025 instead of 2026) and 'Tmm' instead of 'Time'.
• − The composition feels a bit cluttered with the thorns overlapping the bright parchment.

GPT Image 1

• + Successfully adds a thick head of hair
• + Matches the dark color of the beard
• + Maintains the background and main jacket details correctly

• − The hair looks like a separate layer or a wig placed on top
• − Noticeable softening of facial details and skin texture compared to the source
• − The hairline is too sharp and lacks realistic transition

Vidu Q2

• + Excellent hair texture and realistic blending with the existing skin
• + Perfectly preserves facial identity and fine skin details
• + The hairline looks incredibly natural and matches the head shape well

• − None notable for this specific request

GPT Image 1

• + Perfectly clean and legible typography follow the specific prompt layout.
• + Very high-quality PBR materials with soft, clay-like textures.
• + Excellent composition with a professional, balanced 3D isometric look.

• − The flag icon is slightly simplified compared to the rest of the high-quality render.

Vidu Q2

• + Good variety of sushi types on the plate.
• + Adheres well to the requested 45-degree isometric angle.

• − The text rendering is inconsistent, especially around the edges of 'SUSHI'.
• − Visible artifacts and 'hallucinated' shapes around the base of the diorama.
• − The flag icon is positioned awkwardly on a pole growing out of the letter 'N'.

GPT Image 1

• + Excellent caricature style with exaggerated facial features that remain recognizable.
• + Thoughtful integration of all prompt elements including a news desk, dogs, and hockey graphics.
• + Maintains the subject's clothing style and hair color from the source image.

• − The 'TV 13' screen in the background has a slightly distorted box shape.

Vidu Q2

• + Successfully merges a hockey rink with a news set background.
• + Captures the subject's likeliness well in a clean digital illustration style.
• + Includes creative details like paw prints on the news scripts.

• − Anatomical issues with the left hand (extra finger/distorted palm) and right hand (poor grip on microphone).
• − The caricature style is less 'exaggerated' and more like a standard comic book illustration.

GPT Image 1

• + Excellent depiction of god rays and warm golden sunrise lighting
• + Very high fur detail and realistic textures
• + Superior composition with a more focused and intimate 'tumbling' interaction

• − The fox kit has a slightly awkward paw/leg anatomy in the foreground
• − Fewer butterflies compared to the other model

Vidu Q2

• + Includes a large number of colorful butterflies and a dense wildflower meadow
• + Captures a very bright and joyful color palette
• + Good variety of animal poses across the frame

• − Included two golden retriever puppies instead of the requested one
• − Visual quality is less realistic with flatter lighting and less defined fur texture
• − Anatomical errors such as the kitten having three front legs

GPT Image 1

• + Captures the specific 'Studio Ghibli' painterly aesthetic with soft textures and watercolor-like blending.
• + Maintains the warm, nostalgic mood and color palette requested in the prompt.
• + Provides a cohesive artistic transformation while keeping the recognizable pose and characters.

• − The characters' faces are significantly simplified and lose the likeness of the original subjects.
• − The plaid pattern on the man's shirt is largely lost in the painterly texture.

Vidu Q2

• + Preserves the original composition and character details with high accuracy.
• + Successfully translates the scene into a clean anime style with clear line work.
• + Excellent retention of the background elements and the specific patterns on the man's shirt.

• − The style feels more like a generic modern anime or 'manhwa' rather than the specific Ghibli aesthetic.
• − The lighting is flat and lacks the 'dreamy' quality requested in the prompt.

GPT Image 1

• + Excellent source preservation, keeping the light and colors consistent with the original
• + Dynamic hair motion looks natural and integrated
• + Successfully adds subtle motion to the dog's fur and tail

• − The flying leaves are quite small and many look like blurry specs or artifacts
• − The hand holding the leash has become slightly distorted compared to the original

Vidu Q2

• + Strong visual impact with vibrant, colorful flying leaves
• + Hair motion is very clear and well-defined
• + Adds a nice depth of field effect to the leaves in the foreground

• − Leaves look a bit like a flat overlay rather than being part of the environment
• − Changed the color temperature of the entire image, making it warmer/more orange than the source

GPT Image 1

• + Perfect text rendering of 'Caffè Florian' and 'Est. 1720'.
• + Excellent minimalist vector aesthetic with a clear silhouette.
• + Correct inclusion of all requested elements including the cloche and banner.

• − Failed the prompt instruction for a 'light background' by using solid black.
• − Lack of 'cream' tones mentioned in the color palette.

Vidu Q2

• + Followed the background instruction with a subtle texture on a light cream background.
• + Good use of the requested warm brown and cream color palette.

• − Major text hallucinations and typos (e.g., 'Farmiin', 'Esttt', 'Fopli20').
• − Steam is rendered inside the glass cloche, which looks visually confusing.
• − Cluttered composition that lacks the requested minimalism.

GPT Image 1

• + Excellent typography with mostly correct spelling of mission stages and crew names.
• + Strict adherence to the NASA-inspired color palette and flat vector style.
• + Clear, high-contrast layout that functions well as an infographic.

• − One spelling error ('EARLLUNAR' vs Lunar Orbit).
• − The sequence flow is slightly cluttered with icons and text not aligning perfectly in a linear path.

Vidu Q2

• + Clean aesthetic with modern iconography.
• + Good use of spacing and a central vertical symmetry in the layout.

• − Severely garbled text and nonsensical words like 'ALFONCH' and 'Laup 1'.
• − Incorrect number of crew members (5 silhouettes instead of 3).
• − Fails to follow the specific 6-step prompt, mixing up labels and icons.