GPT Image 1.5 OpenAI vs Wan 2.7 Pro Alibaba

GPT Image 1.5

• + Excellent photographic quality with realistic glass reflections and refraction.
• + Perfect lighting consistency between the window light and the shadows on the sphere.
• + Strong aesthetic balance and clean composition.

• − The sphere is quite large, deviating slightly from the 'small blue sphere' description.

Wan 2.7 Pro

• + Successfully includes the window in the frame to justify the light source.
• + Good plant variety that feels natural in the setting.
• + Accurate placement of all requested elements.

• − The glass cube has internal vertical lines that look like structural supports or seams rather than a solid glass object.
• − The red book looks slightly distorted where it meets the glass edge.

GPT Image 1.5

• + Excellent preservation of the specific person's likeness and clothing
• + Accurate integration of the man's hand on the steering wheel
• + Very high quality California coastline background that matches the car's perspective

• − The car is slightly cropped compared to the original composition
• − Minor lighting mismatch between the man and the bright outdoor environment

Wan 2.7 Pro

• + Preserved the entire car from the source image accurately
• + Great lighting and shadow integration on the road

• − Complete failure to use the man from the source image
• − The driver is a generic placeholder with sunglasses

GPT Image 1.5

• + Excellent adherence to the 'candid' and 'imperfect framing' prompt with a tight, ground-level composition.
• + The rain droplets on the man's jacket and the reflections on the ground are highly realistic.
• + The shallow depth of field is well-executed, creating a high-quality cinematic look.

• − The car in the background lacks the requested 'motion blur' and looks like it is parked.
• − The anatomy of the bicycle frame becomes confusing near the rear wheel.

Wan 2.7 Pro

• + Successfully captured more of the urban Japanese environment in the background.
• + The man's skin texture and clothing are naturally rendered without oily AI sheen.

• − The bicycle rendering is physically impossible, with a frame that doesn't connect logically and spokes that are chaotic.
• − The composition feels like a posed portrait rather than the 'candid street photo' requested.
• − Failed to include any motion blur for passing cars.

GPT Image 1.5

• + Exceptional rendering of skin texture, dirt, and realistic scar tissue
• + Intricate engraving and metallic luster on the armor with high-fidelity cloth textures
• + Perfectly executed bokeh sparks and warm lighting that integrates with the subject

• − The beads in the hair are somewhat uniform and less varied in material than they could be

Wan 2.7 Pro

• + Strong composition with visible torches providing context for the lighting
• + Creative interpretation of the braids and colorful beads
• + Good adherence to the 'battle-worn' descriptor with visible facial scars

• − The armor engraving looks slightly flatter and less detailed compared to Model A
• − Leather straps lack the high-detail grit and weathered texture requested in the prompt

GPT Image 1.5

• + Excellent text legibility and correct spelling.
• + Well-organized hierarchy with clear section headers.
• + High-quality, appetizing food photography that looks professional.

• − The layout is a bit generic for a 'modern' design.
• − Missing the 'Mains' category photo grid, focusing only on appetizers and pizza.

Wan 2.7 Pro

• + Beautiful aesthetically pleasing layout with a strong grid system.
• + Sophisticated branding elements including logo, QR code, and social icons.
• + Creative presentation with background props.

• − Text is largely nonsensical and illegible upon close inspection.
• − The food photos contain minor AI artifacts and inconsistencies.
• − Tiny text makes it non-functional as a real menu mockup.

GPT Image 1.5

• + Excellent typography with a fiery, glowing effect that perfectly matches the prompt.
• + Superior textures on the meat and bun create a highly realistic and appetizing look.
• + Dynamic composition with a strong sense of heat and energy.

• − The 'exploded' effect is slightly more compressed than Model B.
• − The bottom bun looks a bit over-charred.

Wan 2.7 Pro

• + Clearer separation of individual burger components in the 'exploded' view.
• + Clean, minimalist background that makes the ingredients pop.

• − Failed to include 'LIMITED TIME ONLY' and '€6.99' in a starburst as requested.
• − The 'MAGIC BURGER' text is plain orange and lacks the requested fiery, glowing effect.
• − The lighting on the ingredients feels a bit artificial compared to the environment.

GPT Image 1.5

• + Excellent chalk texture and smudging effects for a realistic feel.
• + Perfect spelling and completion of the unfinished prompt text.
• + Consistent and authentic handwriting style that looks truly manual.

• − Limited background context compared to the cozy café requested.

Wan 2.7 Pro

• + Beautiful composition including the café interior and framing.
• + Clean, legible text rendering with high contrast.

• − Text appears more like a digital font than natural chalk handwriting.
• − Severe repetition error with 'Grilled Octopus' line appearing twice.
• − Failed to produce the 'elegant cursive' style for the title.

GPT Image 1.5

• + Successfully integrated the man from Image 2 into the scene.
• + Maintained character attributes including sunglasses, scarf, and specific clothing details.
• + Accurately replicated the yellow background and red ottoman from Image 1.

• − Failed to match the extreme 'bent over' torso angle and head tilt of the reference pose.
• − The anatomy of the feet is significantly distorted with too many toes.
• − The right hand contains an extra finger and lacks anatomical consistency.

Wan 2.7 Pro

• + Maintained the exact original image composition and pose.

• − Completely failed the edit instruction by returning a slightly blurred version of Image 1.
• − Failed to include any elements from the character reference (Image 2).

GPT Image 1.5

• + Excellent cinematic lighting and texture on the horse's coat.
• + Rich background detail including a lunar lander and nebula-like starfields.
• + Dynamic composition with kicked-up lunar dust adding to the realism.

• − Failed the negative prompt instruction 'horse on top, not vice versa' by placing the astronaut on top.
• − Anatomical issues with the horse's front legs and hoof attachments.

Wan 2.7 Pro

• + Successfully interpreted the surreal nature of the prompt by letting the horse float in space.
• + Clean, high-resolution rendering with a polished digital art aesthetic.
• + Accurate astronaut suit details and a sense of scale relative to the Earth below.

• − Failed the specific positional instruction 'horse on top, not vice versa'.
• − The floating planets in the background feel a bit cluttered and copy-pasted.

GPT Image 1.5

• + Excellent transfer of the specific outfit components including the coat, scarf, glasses, and watch.
• + Maintains the subject's unique vitiligo patterns and facial features accurately.
• + Perfectly aligns the pose and lighting of the clothing with the original subject.

• − Crop is tighter than the original source image, losing some background context.
• − The skin under the sunglasses doesn't perfectly match the original eye area's texture.

Wan 2.7 Pro

• + Preserves the full composition and framing of the original scene.
• + Maintains the full body pose including the placement of the feet on the sand.

• − Completely failed to use the correct outfit from Image 2, generating a generic gold-patterned blazer instead.
• − Artifacts on the hands such as merging fingers and inconsistent skin patterns.
• − The face is distorted and loses the specific likeness of the person in the source image.

GPT Image 1.5

• + Excellent texture on the capybara's fur and the taxi dashboard.
• + Perfect adherence to the requested composition with the passenger in the background.
• + Highly realistic lighting and cinematic bokeh.

• − The capybara's paws look slightly more like badger or otter paws than true capybara feet.

Wan 2.7 Pro

• + Natural-looking hands on the steering wheel.
• + Clear, high-resolution rendering of the urban background.

• − Failed to place the passenger in the back seat, putting her in the passenger seat instead.
• − The capybara's head is not naturally attached to the body, appearing like a mask or a floating element.
• − The composition feels less like a real interior shot due to the wide side-angle.

GPT Image 1.5

• + Excellent adherence to the 'vintage dark parchment' aesthetic
• + Text is perfectly rendered and integrated into the design
• + Cinematic lighting and moody atmosphere are very effective

• − The thorns and webs border is a bit messy and cluttered

Wan 2.7 Pro

• + Clean, illustrative style with good composition
• + Accurate text rendering for all requested fields
• + Creative border elements like skulls and roses

• − Lacks the requested 'dark parchment' and vintage gothic feel, appearing more like a modern digital illustration
• − The central pumpkin has a strange candle-nose artifact

GPT Image 1.5

• + Excellent texture and density integration with a natural-looking curly style.
• + Maintains very high fidelity to the original facial features and expression.
• + Seamless blend between the sideburns and the new hair.

• − Higher hairline than Model B, bordering on a slightly receded look despite being 'full'.

Wan 2.7 Pro

• + Successfully adds a full head of hair that matches the character's aesthetic.
• + Higher degree of source preservation for the background and jacket textures.
• + Good hair texture and realistic volume.

• − Slightly alters the eye/eyebrow area, making the subject look a bit younger or different than the source.
• − Hair texture feels slightly more 'painted' compared to the coarse realism in Model A.

GPT Image 1.5

• + Excellent PBR materials with realistic textures on the wood and ceramics
• + Follows the diorama base instruction perfectly with a multi-layered square block
• + Clear, bold text rendering and accurate flag placement

• − The scale of the tea set and soy sauce bottle makes the 'miniature' diorama feel a bit crowded

Wan 2.7 Pro

• + Captures the 'soft refined textures' and cartoon aesthetic very effectively
• + Clean, minimalist composition that feels very high-clarity
• + Modern typography and layout

• − Missed the 'diorama base' requirement, placing the plate directly on the floor
• − The flag icon is placed to the right instead of at the top-center as requested

GPT Image 1.5

• + Excellent character likeness preserved while applying the caricature style
• + Very high visual quality with 3D-painted textures
• + Clever integration of hobbies like the dog wearing a hockey helmet

• − The fingers on the hand holding the microphone are slightly warped

Wan 2.7 Pro

• + Successfully incorporates all prompt elements including speech bubbles
• + The facial expression is appropriately exaggerated for the request

• − The art style is more generic 2D clip-art compared to Model A
• − Likeness to the source image is significantly weaker
• − The 'hockey stick' held in the hand is illogical, merging into a microphone stand

GPT Image 1.5

• + Perfectly captures the 'tumbling together' aspect of the prompt with a dynamic and cuddly composition.
• + Superior fur texture and lighting, creating a warm, golden glow with atmospheric god rays.
• + Excellent expressions on all animals that convey a truly joyful and wholesome vibe.

• − The kitten has an extra toe/pad visible on its raised paw.
• − The fox's anatomy is slightly merged with the puppy on the right side.

Wan 2.7 Pro

• + Better individual limb definition and separation between the animals.
• + Realistic inclusion of all requested animals in a full-body action pose.
• + Clean butterfly renders with varied colors.

• − The animals appear more like they are standing next to each other rather than 'tumbling together'.
• − The kitten's facial structure is slightly distorted and less 'cute' than the others.
• − The lighting feels a bit more flat and less atmospheric than the requested '8K masterpiece' glow.

GPT Image 1.5

• + Captures the characteristic Ghibli 'dreamy' lighting and soft color palette perfectly.
• + Successful re-interpretation of the faces into a hand-drawn anime style while keeping the expressions recognizable.
• + Excellent use of soft textures and a warm, nostalgic mood as requested.

• − The foreground character is quite blurry, making it feel less like a sharp illustration and more like a filter.

Wan 2.7 Pro

• + Excellent preservation of the source image's composition and minor details like the plaid pattern on the shirt.
• + Very high clarity and sharpness for a watercolor-style illustration.
• + Accurately conveys the hand-painted texture mentioned in the prompt.

• − The art style is more generic Western watercolor/sketch rather than specific Studio Ghibli anime style.
• − The man's face retains too many realistic features (like stubble detail) to feel like a Ghibli character.

GPT Image 1.5

• + Successfully added a large volume of falling leaves
• + Effectively captures the wind blowing the hair outwards
• + Maintains high color consistency with the original photo

• − The leaf placement looks a bit flat and like a multi-layered overlay
• − Slightly alters the facial features of the woman compared to the source

Wan 2.7 Pro

• + Natural integration of leaves with varied motion blur
• + Hair movement looks realistic and follows a consistent wind direction
• + Excellent preservation of the woman's face and the dog's appearance

• − Fewer leaves than Model A might feel less 'energetic' to some users

GPT Image 1.5

• + Excellent typography with correct capitalization and accent mark
• + Strong vector-style illustration with high-quality stippled shading
• + Centralized and balanced composition for a logo

• − Failed the instruction for a light background, delivering a black one instead
• − The Cloche has a slightly modern/glossy lighting effect rather than purely minimalist vintage

Wan 2.7 Pro

• + Successfully followed the requirement for a light background with subtle texture
• + Exceptional vintage aesthetic with delicate line-art illustration
• + Comprehensive emblem design with extra details like geographical location

• − Spelling error in the primary brand name (Florion instead of Florian)
• − Clutter at the edges of the frame distracts from the central logo emblem

GPT Image 1.5

• + Excellent adherence to the color palette and flat-vector style.
• + Clear, large-scale iconography that is easily readable.
• + Perfectly legible and accurate spelling for all step titles and astronaut names.

• − The layout feels a bit cramped within the individual panels.
• − The Saturn V rocket design is slightly generic and lacks the iconic black/white pattern.

Wan 2.7 Pro

• + Superior infographic layout with a logical vertical flow and mission data.
• + Includes extra details like dates, sites, and mission duration that add to the poster feel.
• + Creative use of minimalist icons that fit the modern vector aesthetic well.

• − Multiple spelling errors in the text (e.g., 'DESCRIPT', 'Tranquilicy', 'minures').
• − The NASA logo in the corner is distorted and contains gibberish.