Wan 2.5 (Preview) Alibaba vs Wan 2.7 Alibaba

Wan 2.5 (Preview)

• + Excellent photographic lighting and depth of field
• + Clean, minimalist composition
• + High-quality textures on the sphere and wood

• − The glass cube is missing its back right vertical edge
• − The book appears to be floating slightly above the glass surfaces

Wan 2.7

• + Strong spatial coherence and realistic glass refraction
• + Highly detailed textures on the book spine and weathered wood
• + Accurate light interaction with the window in the background

• − The placement of the sphere's reflection on the left panel is slightly confusing geometrically

Wan 2.5 (Preview)

• + Excellent preservation of the specific car model and features from the source.
• + High visual quality with realistic motion blur in the wheels and road.
• + Successfully incorporates the subject's dreadlocks and black scarf inside the vehicle.

• − The driver's face is small and lacks detail due to the angle.
• − The palm trees appear a bit repetitive in their placement.

Wan 2.7

• + Beautiful composition of the California coastline and cliffside road.
• + Accurately places the car in a scenic sunset/golden hour lighting environment.

• − Completely fails to preserve the identity of the man, replacing him with a distorted, different person.
• − The car's front grille and headlight details are slightly more muffled compared to the source.

Wan 2.5 (Preview)

• + Excellent shallow depth of field and bokeh realism
• + Rich, saturated colors and beautiful reflections on wet surfaces
• + Stronger 'cinematic' feel requested by the prompt

• − Failed to include passing cars for the motion blur effect
• − Composition is very balanced and centered, lacking the 'imperfect framing' requested

Wan 2.7

• + Successfully captured the 'imperfect framing' and 'candid' street photography style
• + Higher technical accuracy for the environment and background elements
• + More realistic skin texture and clothing details for a candid photo

• − The depth of field is deeper than the 'shallow 50mm' look requested
• − Lacks the motion blur of passing cars requested in the prompt

Wan 2.5 (Preview)

• + Excellent engraving details on the pauldrons
• + High-quality skin texture with realistic dirt and minor bruising
• + Atmospheric warm lighting effects on the armor

• − The beads in the hair look more like metal spheres than decorative beads
• − The background torch has some digital artifacts and looks slightly disconnected

Wan 2.7

• + More natural and varied beadwork in the braids
• + Character looks significantly more 'battle-worn' and experienced
• + Superior depth of field with convincing golden bokeh

• − The engravings on the chest plate are slightly less sharp than Model A's pauldrons
• − One of the leather straps passes through a metal plate in a way that suggests a clipping error

Wan 2.5 (Preview)

• + Strict adherence to the requested grid layout for sections.
• + Follows the request for bold sans-serif fonts and specific section names.
• + High resolution, clean white background with professional lighting on food.

• − Nonsense text and significant spelling errors throughout.
• − Repetitive food imagery (all dishes look very similar with basil toppings).

Wan 2.7

• + Excellent text legibility and realistic branding details including prices and social media icons.
• + Diverse and appetizing food photos that actually correspond to classic menu items.
• + Highly polished, professional composition with lifestyle flat-lay elements.

• − Missing the specific 'Appetizers/Pizza/Mains' grid layout, showing a general grid instead.
• − The background is not purely white as it includes table props.

Wan 2.5 (Preview)

• + Excellent photorealistic texture on the patty and bun.
• + The glowing sauce drip on the title text is creatively executed.
• + Balanced composition with a strong sense of upward motion.

• − The starburst for the price is slightly irregular in shape.

Wan 2.7

• + High-energy composition with great use of liquid splashes and embers.
• + Crisp and clear typography for all three required text elements.
• + Includes extra details like pickles and sesame seeds flying in the air.

• − The bun looks slightly less realistic and more like a 3D render than Model A.
• − The 'LIMITED TIME ONLY' text feels a bit crowded at the bottom.

Wan 2.5 (Preview)

• + Text has a very convincing chalk texture with authentic 'dustiness' and smudges.
• + Handwriting looks truly manual with natural slants and variations in stroke weight.
• + Accurately rendered all the requested text and prices with high legibility.

• − The title is in a blocky style rather than the 'elegant cursive' requested.
• − The perspective is slightly angled, making some text less prominent in the composition.

Wan 2.7

• + Perfectly centered and balanced composition.
• + Followed the text content prompt accurately for the menu items.
• + The background environment (cafe wall) is aesthetically pleasing.

• − The text looks like a digital font rather than natural chalk handwriting.
• − The 'chalk' texture is too uniform and lacks the physical grainy characteristics shown in Image A.
• − Failed the 'elegant cursive' requirement for the title.

Wan 2.5 (Preview)

• + Excellent dynamic lighting with cinematic lens flares.
• + Vibrant color palette and deep space contrast.
• + Detailed textures on the horse's coat and astronaut's suit.

• − Failed the specific spatial instruction (the astronaut is riding the horse, not vice-versa).
• − Anatomical issues with the horse's front-left leg and floating stirrup.

Wan 2.7

• + Natural and balanced composition with a clear view of the Earth.
• + Good rendition of the horse's mane and bridle hardware.
• + Sharp clarity throughout the entire frame.

• − Completely ignored the negative constraint/instruction for the horse to be on top.
• − The horse's legs are disproportionately long and thin.
• − Static posing lacks the 'surreal' or 'cinematic' energy requested.

Wan 2.5 (Preview)

• + Successfully replicates the exact clothing, scarf, and accessories from Image 2.
• + Maintains a realistic lighting match between the person and the beach background.

• − Completely fails to preserve the identity, face, and hair of the person in Image 1, replacing them with the person from Image 2.
• − Crops the image to a square format, losing parts of the original composition.

Wan 2.7

• + Perfectly preserves the person's exact face, hair, and unique skin patterns from Image 1.
• + Successfully adapts the pose and body shape while maintaining the background and original aspect ratio.

• − Fails to use the specific clothing from Image 2, instead generating a generic 'elaborate' outfit.
• − The generated hands have anatomical issues and strange skin blending.

Wan 2.5 (Preview)

• + Excellent photographic quality with realistic rain droplets and lighting.
• + Correct placement of the passenger in the back seat as requested.
• + Highly detailed rendering of the capybara's fur and the taxi interior.

Wan 2.7

• + Sharp rendering of the capybara's coat and human features.
• + Clear view of the city streets through the window.

• − Failed to place the passenger in the back seat; she is sitting in the front passenger seat.
• − Composition feels slightly cramped with the capybara's head overlapping the door frame.
• − The paws on the steering wheel look anatomically strange and lack the weight found in the other image.

Wan 2.5 (Preview)

• + Excellent cinematic lighting within the central jack-o-lantern and backdrop.
• + High level of texture detail on the pumpkin and the twisted tree.
• + Successful integration of all requested text with perfect spelling.

• − The main title text is cut off slightly by the top and side borders.
• − The composition feels a bit cramped with the large text overlapping the scenery.

Wan 2.7

• + Perfectly balanced layout that feels like a real vintage poster.
• + Crisp, legible text with elegant gothic font choice.
• + Highly detailed and decorative border that incorporates webs, thorns, and skulls beautifully.

• − The lighting is less cinematic and more illustrative/flat compared to Model A.
• − Added extra elements like 'Est. 1847' and 'Dress code' which were not in the prompt.

Wan 2.5 (Preview)

• + Successfully added a thick head of hair with realistic volume.
• + Maintains excellent preservation of background and clothing.
• + The lighting on the hair matches the desert environment well.

• − The hairline on the forehead looks slightly artificial and rounded.
• − The hair texture is a bit too soft or smoothed over in some areas.

Wan 2.7

• + Exceptional hair texture and realism, showing individual strands and natural variation.
• + Excellent hairline integration that looks like it naturally grows from the scalp.
• + Near-perfect source preservation with virtually no changes to the face or environment.

• − The hair volume is slightly more conservative compared to the 'full, thick' request in model A.

Wan 2.5 (Preview)

• + Excellent typography rendering with clean, clear fonts
• + Beautiful soft 3D lighting and material textures
• + Perfectly centered and clean composition

• − Simple single-sushi composition might be considered 'too minimal' for a diorama

Wan 2.7

• + Successfully captures the 45° isometric perspective and diorama base
• + Offers more visual variety with multiple types of sushi and garnishes
• + Accurate rendering of the flag icon and text layout

• − The text has minor aliasing and a distracting white outline
• − Some minor artifacting on the chopsticks and soy sauce dish

Wan 2.5 (Preview)

• + Successfully incorporates all elements: hockey, dogs, and TV anchor desk.
• + Maintains the subject's denim outfit from the source image.
• + Clean, classic caricature style with a balanced composition.

• − The facial resemblance to the source image is somewhat generic.
• − The screen in the background shows a hockey player on a grass-like field instead of ice.

Wan 2.7

• + Captures the subject's facial features and eye color more accurately than Model A.
• + Highly creative and humorous use of space, including a tiny hockey rink and dogs in helmets.
• + Maintains the 'selfie' arm perspective from the original photo.

• − Includes some text spelling errors ('Rolee' instead of 'Rule').
• − The background is slightly more cluttered compared to the clean layout of Model A.

Wan 2.5 (Preview)

• + Excellent depiction of motion and energetic interaction
• + Vibrant colors and strong lighting effects
• + Distinct and cute facial expressions on all characters

• − Physical anatomy on the fox is slightly distorted
• − Dew drops appear as large floating orbs rather than being on the grass
• − The fox's eyes appear somewhat artificial and glowing

Wan 2.7

• + Great fur textures and more realistic lighting transitions
• + Better integration of the animals into the meadow
• + More naturalistic anatomy for the rabbit and fox

• − The kitten has a slightly strange expression and pose
• − Lower energy in the composition compared to the other image
• − The light rays are a bit more diffused and less 'defined' than requested

Wan 2.5 (Preview)

• + Captures the iconic clean anime line art style specific to Studio Ghibli films.
• + Excellent facial expression mapping that translates the real-life meme to an anime context perfectly.
• + Adds whimsical elements like floating leaves and sparkles that enhance the dreamy mood.

• − Faces are slightly generic 'modern anime' rather than strictly the traditional round Ghibli aesthetic.

Wan 2.7

• + Strong hand-painted watercolor texture throughout the image.
• + Preserves the physical likeness of the original actors more closely than Model A.

• − The style feels more like a generic European watercolor illustration than a Studio Ghibli anime style.
• − The lighting is flat and lacks the 'glow' or 'wonder' requested in the prompt.

Wan 2.5 (Preview)

• + Excellent wind effect on hair
• + Successfully added motion blur to the dog's tail
• + Consistent facial features and clothing preservation

• − The added leaves look like flat digital stickers rather than natural elements
• − The green color of the leaves is overly saturated and neon

Wan 2.7

• + Natural-looking leaves with realistic colors and textures
• + Leaves show a good sense of depth and motion
• + Excellent preservation of the subject and background

• − Hair movement is less dynamic than in the other model
• − Motion blur on the dog is subtle compared to the other model

Wan 2.5 (Preview)

• + Excellent typography rendering with perfect spelling and accent marks.
• + Clean vector-style composition that feels like a professional logo.
• + Great use of the background texture to enhance the vintage feel.

• − The steam is a bit thick and stylized compared to the rest of the logo.

Wan 2.7

• + Strong emblem layout with a coherent circular frame.
• + Includes nice decorative elements like the stars and wheat stalks.
• + Accurate adherence to the color palette.

• − The spelling 'Florion' is incorrect, missing the 'a'.
• − The cloche is transparent, which looks a bit awkward with the steam placement behind it.
• − The layout is slightly cluttered for a 'minimalist' request.

Wan 2.5 (Preview)

• + Excellent adherence to the 'flat-vector' and 'crisp lines' instruction.
• + Clean and legible text for the main headers.
• + Strong visual storytelling through connected icons.

• − Missed the labels for 'Descent' and 'Landing' next to the actual graphics.
• − The astronaut portraits are a bit jarring compared to the flat vector style of the rest of the image.

Wan 2.7

• + Perfectly organized vertical flow that matches the chronological order of the prompt.
• + Included much more detailed technical text and data that feels like a real infographic.
• + Excellent text rendering for 1960s-era mission names and details.

• − Contains a misspelling ('DESCRIPT' instead of 'DESCENT').
• − The icons are much smaller and less impactful than those in Model A.