GPT Image 1 Mini OpenAI vs Wan 2.6 Alibaba

GPT Image 1 Mini

• + Perfectly matches all spatial requirements including sphere inside and book on top.
• + Clean and minimalist composition with very realistic lighting.
• + Glass edges are sharp and accurately rendered.

• − The sphere is quite small compared to the scale of the cube.
• − The background plant is very blurred, making it harder to see 'through' the glass as requested.

Wan 2.6

• + Excellent texture on the red book and wooden table.
• + Lighting through the window is very dynamic and natural.
• + The plant is clearly positioned behind and visible through the glass.

• − The glass cube has strange internal geometry with a visible divider in the middle.
• − The blue sphere is quite large, stretching the 'small' descriptor in the prompt.

GPT Image 1 Mini

• + Excellent preservation of the man's facial features and specific hairstyle.
• + High fidelity to the original plaid coat and black snood.
• + Natural lighting and motion blur that creates a cohesive scene.

• − The scale of the man inside the car is slightly too large, making the car look small.

Wan 2.6

• + Very accurate representation of the California coastline with palm trees and cliffs.
• + Good composition with the car and environment.
• + Maintains the car's branding and design elements well.

• − The man's facial likeness is significantly altered, losing the specific features of the source image.
• − The man's hair texture and volume are less accurate than the first model.

GPT Image 1 Mini

• + Excellent skin texture and hyper-realistic facial details
• + Accurate 50mm shallow depth of field effect
• + Higher overall visual fidelity in the foreground

• − The red bicycle is partially cropped, reducing narrative impact
• − The light rain is barely visible in the atmosphere

Wan 2.6

• + Better 'candid' environmental composition showing the full scene
• + Strong execution of wet pavement reflections and light rain visible on clothing
• + Dynamic background with motion blur from passing cars

• − Noticeable anatomy issues with the hands appearing distorted
• − The raindrops on the jacket appear as static white dots rather than natural streaks
• − Lower level of fine facial detail compared to Image A

GPT Image 1 Mini

• + Excellent depiction of warm torchlight reflecting off the metal surfaces.
• + Highly detailed and realistic skin texture with subtle scarring and dirt.
• + Intricate engraving on the plate armor that looks consistent and weighted.

• − The braids are very thin and subtle, almost blending into the hair.
• − Beads in the hair are missing or not clearly visible as requested.

Wan 2.6

• + Strong adherence to the 'beads' and 'braids' prompt with clearly visible decorations.
• + Excellent texture on the leather straps and frayed cloth underlayer.
• + Very lifelike, expressive eyes that convey the 'battle-worn' theme well.

• − The grime on the face looks somewhat like digital smudges rather than integrated dirt.
• − Some of the floating sparks/bokeh in the foreground look a bit artificial and distracting.

GPT Image 1 Mini

• + Excellent text legibility and clean sans-serif typography.
• + Perfect alignment between section headings and corresponding food types in the grid.
• + High-quality, isolated food photography that fits the minimalist aesthetic.

• − The menu content is empty, showing only headings without item descriptions or prices.
• − The layout is a bit overly simplistic, bordering on a template rather than a finished design.

Wan 2.6

• + Successfully incorporates vibrant color accents as requested in the corners and lines.
• + Includes realistic menu elements like prices, item names, and descriptions.
• + Features a greater variety of food photography within the grid.

• − Contains significant gibberish text and spelling errors (e.g., 'Resstaurantsr2eher').
• − Poor organization where 'Pizza' appears in the grid but the section header is below the photos, and 'Appetizers' contains photos of pizza.

GPT Image 1 Mini

• + Excellent and clean text rendering for all elements
• + Highly professional studio-style lighting on the burger components
• + Perfect adherence to the starburst and fiery glowing text requirements

• − Static composition that lacks the 'exploded' motion requested
• − The background is slightly too plain compared to the fiery request

Wan 2.6

• + Dynamic 'exploded' composition with a true sense of motion and flying ingredients
• + Very detailed fiery background with smoke and embers
• + High-quality textures on the meat and sauce

• − The 'MAGIC BURGER' text has slight irregularities in the 'B' and 'A'
• − The starburst is a generic cartoon graphic rather than the requested fiery glowing effect

GPT Image 1 Mini

• + Excellent text legibility and accuracy with zero spelling errors.
• + Stable and clean composition.
• + Consistent chalk texture across all lettering.

• − The 'cursive' request for the title was not followed, as it remains blocky.
• − The text looks slightly too uniform, bordering on a digital font look rather than true organic handwriting.

Wan 2.6

• + Captures the 'elegant cursive' and 'handwritten' request much more authentically.
• + The texture of the board with chalk dust and smudges provides superior realism and atmosphere.
• + Good adherence to the slanted handwriting requested in the prompt.

• − Small spelling error in the date ('APRIL' is misspelled as 'APRIL' but the 'L' is mangled and there is an extra 'I').
• − The composition is slightly tighter at the edges.

GPT Image 1 Mini

• + Excellent cinematic lighting and texture on the space suit.
• + Crisp resolution and realistic rendering of the horse's musculature.

• − Failed the negative constraint entirely; the astronaut is riding the horse.
• − The composition is a bit standard and lacks the requested surrealism.

Wan 2.6

• + High level of detail in the astronaut's gear and the cosmic background.
• + Dynamic composition with vibrant colors and lighting.

• − Failed the specific spatial instruction for the horse to be 'on top'.
• − The prompt requested surrealism, but this remains a literal interpretation of an astronaut riding a horse.

GPT Image 1 Mini

• + Excellent adherence to the request for the 'exact elaborate outfit' including the watch and belt.
• + Successfully maintains the vitiligo pattern on the visible hands.
• + Provides a full-body view that demonstrates logical garment fitting.

• − Significantly alters the subject's facial features and hair, failing the 'completely unchanged' instruction.
• − The subject's skin tone appears more muted and less vibrant than the original source.

Wan 2.6

• + Near-perfect preservation of the source person's face, skin texture, hair, and sand on the cheek.
• + Accurately places the scarf and coat while matching the perspective of the original pose.
• + Background remains highly consistent with the original source image.

• − Missed several components of the outfit such as the watch, belt, and shoes by cropping the image.
• − Added sunglasses that were not part of the person's face in Image 1 and look different from the ones in Image 2.

GPT Image 1 Mini

• + Excellent photorealism and cinematic lighting
• + High quality texture detail on the capybara's fur
• + Accurate moody atmosphere for a night taxi ride

• − The passenger is seated correctly in the back but slightly out of focus
• − Only one paw is clearly visible and positioned on the steering wheel

Wan 2.6

• + Follows the instruction for both front paws on the steering wheel better
• + Includes vibrant New York street lights in the background
• + Captures the bored expression of the businesswoman well

• − Layout error where the passenger appears to be in the front seat next to the driver rather than the back seat
• − The capybara's hat looks more like a police officer hat than a taxi driver cap
• − Visual artifacts on the car roof

GPT Image 1 Mini

• + Excellent text legibility and alignment
• + Strong parchment texture that feels cohesive with the vintage theme
• + Balanced layout with clear hierarchy

• − Very dark color palette limits the visibility of the twisted trees
• − The 'webs and thorns' border is less intricate than requested

Wan 2.6

• + Ornate and detailed border with thorns and cobwebs
• + Highly cinematic lighting with a vibrant blue sky and orange glow contrast
• + Great attention to detail in the twisted tree textures

• − Text rendering is slightly inconsistent, with 'a' looking like 'o' in 'to'
• − The text lacks the perfect alignment of Model A

GPT Image 1 Mini

• + Successfully added a thick head of hair with realistic texture
• + Maintains the lighting and environment of the original photo

• − Significantly altered the facial features, making the man unrecognizable compared to the source image
• − Changed the style of the glasses and smoothed out the skin texture

Wan 2.6

• + Excellent source preservation, keeping the facial features and glasses almost identical to the original
• + Hair density and texture look very natural and well-integrated

• − The hairline on the forehead is slightly too low, creating a slightly compressed facial appearance

GPT Image 1 Mini

• + Excellent soft 3D textures and smooth lighting
• + Very clean and bold typography
• + Subtle but effective wood grain texture

• − Perspective is slightly lower than the requested 45° isometric angle
• − Chopsticks were not explicitly requested, though they fit the theme

Wan 2.6

• + Perfect adherence to the 45° isometric diorama request
• + Stronger miniature 'toy' aesthetic
• + Correct placement of 'JAPAN' above 'SUSHI'

• − The flag is placed to the left instead of being small/after text as in most clean graphic designs
• − The gray base feels slightly heavy compared to the soft aesthetic requested

GPT Image 1 Mini

• + Excellent preservation of the subject's facial features and denim shirt within a caricature style.
• + Perfectly follows the caricature art style with hand-drawn textures and exaggerated proportions.
• + Includes all requested elements: anchor profession, dog, and hockey gear.

• − The hand holding the microphone is a bit simplified and claw-like.

Wan 2.6

• + Creative inclusion of multiple dogs and a complete TV studio set.
• + Clean, modern vector illustration style with vibrant colors.
• + Incorporates a hockey jersey on the dog for extra thematic flair.

• − The facial resemblance to the source image is significantly lower than Model A.
• − The caricature is less 'exaggerated' in the traditional sense and more of a generic cartoon character.

GPT Image 1 Mini

• + Excellent fur texture and sharpness on all four animals.
• + Better composition with a clear sense of 'tumbling' and movement.
• + Natural lighting that enhances the golden retriever and kitten fur.

• − The fox's anatomy is slightly stiff compared to the others.
• − Butterfly wings have very basic patterns.

Wan 2.6

• + Stronger 'god rays' and atmosphere with beautiful dew sparkles.
• + Highly detailed butterfly wing patterns and more complex foreground floral elements.
• + Correct animal count and species depiction.

• − The golden retriever's face and paws look slightly distorted/melted into the grass.
• − The kitten's pose and anatomy look less natural compared to the others.
• − Significant blurring on the white bunny's face.

GPT Image 1 Mini

• + Successfully captures the specific facial expressions from the meme in an illustrative style.
• + Excellent use of textured, colored-pencil aesthetic that feels hand-painted.
• + Preserves the composition and character poses almost perfectly.

• − The color palette is a bit overly yellow/warm, losing some of the 'pastel' variety requested.
• − The background architectural details are very blurred compared to the original.

Wan 2.6

• + Captures the Studio Ghibli watercolor aesthetic more accurately with clean line work and soft washes.
• + Maintains the likeness of the original subjects very well while translating them to anime.
• + Includes 'dreamy' lighting elements like soft bokeh/sparkles that enhance the mood.

• − The man's facial expression is slightly less 'guilty' or exaggerated than the original meme.
• − The background characters are a bit simplified.

GPT Image 1 Mini

• + Successfully added a large volume of flying leaves
• + Clearly modified the hair to flow outward
• + High level of energy conveyed through the environment

• − Significantly altered the woman's facial features and clothing details
• − Poor source preservation as it essentially redrew the characters
• − Lost the original dog's specific pose and look

Wan 2.6

• + Excellent source preservation of the woman's face and original clothing
• + Accurate hair motion that stays consistent with the original style
• + Correctly added the requested flying leaves while keeping the background intact

• − The green leaves look slightly artificial and flat compared to the scene
• − Motion is a bit more subtle than the 'energetic' prompt might suggest

GPT Image 1 Mini

• + Excellent typography with perfect spelling and accents.
• + High-contrast, clean vector style that looks professional.
• + Accurate representation of the requested banner and text elements.

• − Ignored the 'light background' instruction, opting for black.
• − The texture is very subtle, almost appearing as noise rather than a vintage paper texture.

Wan 2.6

• + Followed the 'light background' and 'subtle texture' instructions perfectly.
• + Clean, modern interpretation of a minimalist logo.
• + Good spatial balance between the icon and text.

• − The banner for 'Est. 1720' is unusually small and poorly integrated.
• − The 'Est. 1720' text is slightly warped and less legible than Model A.

GPT Image 1 Mini

• + Excellent adherence to the infographic structure, including all 6 requested steps.
• + Clean flat-vector style with a perfect NASA-inspired color palette.
• + Perfect text rendering for all step titles and the crew section.

• − The translunar trajectory line is a bit messy and over-loopy.
• − Minor cropping at the bottom of the image.

Wan 2.6

• + High resolution with a clean aesthetic.
• + Successfully included the names of the three crew members.

• − Completely failed to include the requested 6-step infographic content.
• − The background looks like a fabric texture rather than a vector graphic.
• − Composition is mostly empty space.