GPT Image 1 Mini OpenAI vs Wan 2.5 (Preview) Alibaba

GPT Image 1 Mini

• + Excellent adherence to the 'small blue sphere' prompt requirement
• + Clean and minimalist composition with realistic glass edge rendering
• + Soft, diffuse lighting that creates a calm atmosphere

• − The glass cube is depicted as a hollow frame/vessel without a top face, appearing physically impossible with the book floating on thin edges

Wan 2.5 (Preview)

• + Dynamic lighting with attractive shadows and light rays
• + Higher level of detail in the textures of the book and plant
• + Realistic glass reflections on the bottom surface of the cube

• − The sphere is quite large, ignoring the 'small' descriptor in the prompt
• − Noticeable dust/particle artifacts in the air that may look like noise

GPT Image 1 Mini

• + Excellent preservation of the man's identity, including his hairstyle and specific plaid coat/scarf.
• + Realistic composition and lighting that matches the outdoor sunset setting.
• + Accurate car model details and interior leather matching the source.

• − The steering wheel placement is slightly awkward in relation to his hands.

Wan 2.5 (Preview)

• + Successfully places the car on a California-style coastline with palm trees.
• + Maintain the motion blur on the road and wheels for a 'driving' feel.
• + The car's exterior design is kept very close to the original image.

• − The man is barely visible and loses almost all specific features from the source image.
• − The man's hair and facial features are generic compared to Image A.

GPT Image 1 Mini

• + Excellent skin texture and hyper-realistic facial details
• + Authentic shallow depth of field with 50mm lens look
• + Effective use of subtle rain droplets and muted cinematic color grading

• − The bike anatomy is slightly incoherent around the wheel and chain guard
• − The car in the background is static rather than showing motion blur

Wan 2.5 (Preview)

• + Outstanding reflections on the wet pavement
• + Dynamic wide shot with good environmental storytelling
• + Excellent prompt adherence regarding the rain intensity and street setting

• − The man's feet and shoes look distorted and poorly integrated with the ground
• − The mechanical parts of the bike and tools are a cluttered mess of artifacts

GPT Image 1 Mini

• + Excellent atmospheric lighting and skin texture
• + High level of intricate detail in the plate armor engravings
• + Strong emotional depth in the subject's expression

• − The braids are very subtle and lack the requested beads
• − Leather and cloth textures are somewhat obscured by the dark composition

Wan 2.5 (Preview)

• + Explicitly includes the beads in the braids as requested
• + Superb texture on the leather straps and frayed cloth underlayer
• + Dynamic lighting with a visible torch and high-contrast armor reflections

• − The character looks slightly young for a 'battle-worn' description
• − The armor plating on the far shoulder has some minor geometric inconsistencies

GPT Image 1 Mini

• + Excellent text readability with perfect spelling of headers.
• + Consistent grid layout that aligns images with categories.
• + High image quality for the food items which look appetizing and realistic.

• − Lacks any descriptive menu text or pricing, making it look like a template.
• − Very basic composition that feels almost too empty on the left side.

Wan 2.5 (Preview)

• + Features a more complex and professional menu layout with item descriptions.
• + Includes a variety of colorful food photography that matches the prompt's 'vibrant accents'.
• + Good use of color-coded dividers for different sections.

• − Significant text rendering issues with 'gibberish' descriptions and titles.
• − Several food items look repetitive (the same basil garnish used on almost every plate).

GPT Image 1 Mini

• + Excellent typography rendering with consistent fiery glow effect.
• + Clean, professional layout that balances the text and burger perfectly.
• + Highly realistic food textures, particularly the burger bun and patty.

• − The 'exploded' effect is a bit static and vertical compared to the requested dynamic motion.
• − The background is somewhat empty apart from the glowing embers.

Wan 2.5 (Preview)

• + Strong sense of dynamic motion with ingredients angled and scattered in mid-air.
• + Creative fiery starburst design that incorporates literal flames.
• + Good incorporation of smoke and flying sauce droplets to enhance the 'magic' theme.

• − One slice of tomato and some lettuce leaves appear to be floating somewhat unnaturally thin or poorly lit.
• − The text 'LIMITED TIME ONLY' is slightly warped to follow the curve of the main title.

GPT Image 1 Mini

• + Excellent text legibility and spelling accuracy for all requested items.
• + The chalk texture is very realistic with convincing grain and pressure variations.
• + Perfect adherence to the requested items and pricing.

• − The heading is in print/architectural style rather than the 'elegant cursive' requested.
• − The overall composition is very flat and lacks the 'cozy café' environmental context.

Wan 2.5 (Preview)

• + Successful use of cursive handwriting for the text as requested.
• + Good environmental context showing a café interior in the background.
• + Dynamic composition with natural-looking chalk smudges and smears for realism.

• − Significant text errors including redundant pricing and mangled words in the footer.
• − Failed to include 'Herbs' in the Grilled Octopus line.
• − The handwriting style is slightly more inconsistent in legibility compared to Model A.

GPT Image 1 Mini

• + Excellent moody, cinematic lighting that blends the horse and astronaut into the environment.
• + Strong anatomical consistency for the horse.
• + Captures a mysterious, surreal atmosphere.

• − Failed the specific spatial instruction for the horse to be 'on top' of the astronaut.
• − Lower dynamic range compared to Model B.

Wan 2.5 (Preview)

• + Vibrant colors and high-resolution details in the nebula and planet.
• + Sharp rendering of the space suit textures.
• + Dynamic sense of motion with the dust and hair.

• − Failed the negative constraint/specific instruction for the horse to be 'on top' of the astronaut.
• − The horse's front legs have slightly awkward posing near the hooves.

GPT Image 1 Mini

• + Successfully preserved the unique vitiligo skin patterns from the base person.
• + Correctly transferred the full outfit including the plaid scarf, pea coat, jeans, and watch.
• + Maintained the overall composition and background of the original beach scene.

• − Altered the person's face significantly, making them look older and changing features.
• − Lost the sand textures on the person's skin present in the source image.

Wan 2.5 (Preview)

• + Excellent replication of the outfit details including the sunglasses and specific scarf pattern.
• + Maintained the exact background and lighting of the first image.

• − Completely failed to use the person from Image 1 as the base, using the face and skin of the person from Image 2 instead.
• − Ignored the vitiligo pattern which was the defining feature of the base person.

GPT Image 1 Mini

• + Excellent photorealistic lighting and bokeh effects.
• + High-quality fur texture and realistic anatomy of the capybara paw.
• + Moody, professional cinematic atmosphere that fits the 'night' prompt.

• − Only one paw is visible on the steering wheel.
• − The scene is a bit dark, making it harder to see the yellow taxi details.

Wan 2.5 (Preview)

• + Successfully shows both front paws on the steering wheel.
• + Clearly shows more of the 'Yellow Taxi' environment, including the external lights and signs.
• + Great character expressions for both the capybara and the businesswoman.

• − The paws look slightly like human fingers or claws rather than natural capybara feet.
• − Lower lighting realism compared to Model A, with some over-saturated elements.

GPT Image 1 Mini

• + Excellent font legibility and clean text rendering.
• + Highly cohesive dark parchment aesthetic that fits the 'vintage gothic' prompt.
• + Effective use of subtle atmospheric depth with the trees and bats.

• − The border of thorns and webs is very faint and lacks detail.
• − The lighting on the pumpkin is a bit flat compared to the other model.

Wan 2.5 (Preview)

• + Dynamic and high-quality illustration of the jack-o-lantern with fiery eyes.
• + Border elements like thorns and webs are very detailed and prominent as requested.
• + Strong cinematic lighting and color contrast between the blue sky and orange pumpkin.

• − Text rendering is slightly inconsistent, with a slight misalignment in the top title.
• − The overall composition feels a bit cluttered with the mix of illustrative styles.

GPT Image 1 Mini

• + Successfully added a thick head of hair as requested.
• + Maintains the general color palette of the original.

• − Failed to preserve facial features, significantly altering the subject's nose, eyes, and brow structure.
• − The hair texture looks somewhat blocky and lacks fine-grained realism at the edges.
• − Altered the position and shape of the glasses.

Wan 2.5 (Preview)

• + Excellent source preservation, keeping the facial features and expression nearly identical to the original.
• + High-quality realistic hair texture and natural integration with the sideburns.
• + Preserved the background and clothing details perfectly.

• − The hairline transition on the forehead is slightly sharp, though still realistic.

GPT Image 1 Mini

• + Excellent adherence to the 45-degree isometric perspective.
• + Beautiful text rendering and clean layout.
• + The textures perfectly capture the requested soft refined 3D cartoon style.

• − The rice grains are slightly less defined compared to the other model.

Wan 2.5 (Preview)

• + High clarity on the individual rice grains.
• + Accurate text rendering and placement.
• + Soft, pleasing lighting.

• − Failed to capture the requested 45-degree isometric perspective, opting for a lower eye-level view.
• − The composition feels slightly more generic and less like a miniature diorama.

GPT Image 1 Mini

• + Excellent character likeness to the source person within the caricature style.
• + Charming hand-drawn pencil/crayon aesthetic that feels authentic to a street caricature.
• + Perfectly integrates the hockey puck and stick alongside the news desk theme.

• − The 'NEWS' text is a bit generic compared to a full studio environment.

Wan 2.5 (Preview)

• + Creative use of a hockey jersey on one of the dogs and a game on the monitor.
• + High-contrast digital vector style that is very clean.
• + Clearly incorporates all requested elements in a vibrant composition.

• − The facial likeness is significantly weaker than Model A, appearing more generic.
• − The hand holding the microphone is poorly rendered with an incorrect finger count/structure.
• − The large white border and floating composition feel less like a cohesive scene.

GPT Image 1 Mini

• + Excellent anatomical realism for all four animals.
• + Beautifully soft, naturalistic lighting that feels cohesive with the scene.
• + The fox kit has a very realistic, earthy appearance.

• − The 'god rays' are a bit subtle compared to the other model.

Wan 2.5 (Preview)

• + Stronger adherence to 'god rays' and 'dew sparkles' with very bright lighting.
• + High contrast and vibrant colors that make the subjects pop.

• − The fox's eyes appear unnaturally blue/glowing, almost alien.
• − The kitten's proportions feel slightly off, with a very long neck and small head.
• − The dew drops look more like floating glass spheres than natural dew.

GPT Image 1 Mini

• + Successfully captures a hand-painted, soft texture reminiscent of colored pencils or pastels.
• + Maintains the warm, nostalgic color palette requested.
• + Strongly preserves the specific facial expressions from the original meme.

• − The style feels more like a children's book illustration than the specific Studio Ghibli anime aesthetic.
• − The colors are a bit washed out and lack contrast.

Wan 2.5 (Preview)

• + Perfectly captures the Studio Ghibli character design style and cel-shading.
• + Effectively uses dreamy lighting and floating leaves to enhance the mood.
• + Maintains the exact composition of the source image while translating it into high-quality anime art.

• − The facial expressions are slightly softened compared to the intensity of the original meme.

GPT Image 1 Mini

• + Strong prompt adherence regarding the flying leaves
• + Significant hair movement creates a clear sense of dynamic motion

• − Fails to preserve the original subject's face, making it look like a different person
• − Alters background elements like the bridge significantly

Wan 2.5 (Preview)

• + Excellent source preservation, maintaining the woman's face and original features
• + Realistic hair motion that looks integrated into the original scene
• + Uses green leaves that match the existing foliage colors

• − The flying leaves look somewhat like floating digital elements rather than blowing naturally

GPT Image 1 Mini

• + Excellent typography rendering with correct accent marks
• + Clearer texture within the gold/brown elements
• + Good use of the Est. 1720 banner as requested

• − Completely ignored the request for a 'light background'
• − The black background creates high contrast that feels more modern than 'vintage minimalist'

Wan 2.5 (Preview)

• + Adhered perfectly to the 'light background' request
• + Superior vector emblem style with clean lines
• + Excellent paper texture and vintage atmosphere

• − The accent mark on 'Caffè' is slightly misplaced/oversized compared to standard typography
• − The cloche handle is slightly off-center

GPT Image 1 Mini

• + Excellent text legibility and spelling for all labels.
• + Clean, consistent vector icon style that matches the requested flat aesthetic.

• − The 'Translunar' icon is an abstract scribble that doesn't clearly represent a trajectory.
• − The layout is somewhat fragmented and lacks a cohesive background.

Wan 2.5 (Preview)

• + Sophisticated composition with a dark navy background that feels more space-themed.
• + Includes clever details like the crew portraits with names and a 'Tranquility' marker.

• − Failed to provide separate icons or numbered steps for Descent and Landing.
• − The 'Translunar' label is placed next to a large moon, which is confusing within the sequence of the infographic.