Qwen Image Alibaba vs Vidu Q2 ShengShu Technology

Qwen Image

• + Excellent photographic quality with soft, realistic lighting.
• + Clean and balanced composition.
• + Follows all spatial instructions including the sphere placement and book.

• − The plant is blurred in the background and not very visible through the glass as requested.

Vidu Q2

• + Successfully captures the plant being visible through the glass cube.
• + Realistic wood texture and lighting shadows.
• + Accurate object placement.

• − The book and sphere have a slightly more 'rendered' look compared to the photo-realism of Model A.
• − The glass cube has some thickness inconsistencies on the top edge.

Qwen Image

• + Excellent full-body composition that captures the environment
• + Realistic motion blur on the background vehicles
• + Effective use of shallow depth of field

• − Anatomy errors including a three-legged bicycle stand and floating pedal
• − Faces and hands are a bit soft and lacking fine texture

Vidu Q2

• + Highly detailed 'imperfect framing' that feels like a true candid close-up
• + Superb skin texture on the hands and face
• + Complex mechanical details on the bicycle chain and frame

• − Distorted hands with anatomical blending issues
• − Background car is static despite the request for motion blur

Qwen Image

• + Excellent adherence to the 'beads in hair' prompt with colorful details.
• + Strong atmospheric lighting with visible torchlight and stylized sparks.
• + Intricate engraving details on the armor and clear texture on the underlayer.

• − The facial scars look somewhat like artificial paint or fresh cuts rather than 'faint scars'.
• − The torch in the foreground creates some visual clutter.

Vidu Q2

• + Very realistic skin textures and more natural-looking faint scars.
• + Sophisticated armor engraving and better structural consistency of the chest plate.
• + Cleaner overall composition with a more effective shallow depth of field.

• − The 'braided with small beads' instruction is barely followed, with very few beads visible.
• − The bokeh effects in the background are less dynamic than requested.

Qwen Image

• + Strong minimalist aesthetic with a very clean white background.
• + Excellent font choice that matches the 'bold sans-serif' prompt perfectly.
• + The image grid is perfectly aligned and creates a high-quality professional look.

• − Text contains frequent gibberish and typos like 'Pizzaurant'.
• − The food variety is limited, with several images looking very similar.

Vidu Q2

• + Features a wider variety of food photos and more comprehensive menu content.
• + Follows the multi-section categorization requested in the prompt.
• + Contains vibrant accents and more complex design elements.

• − The layout is cluttered and fails the 'minimalist' requirement.
• − Text rendering is very poor with extreme distortion and illegible characters.
• − Some food items look messy or unappetizing due to low-quality AI generation.

Qwen Image

• + Excellent typography with a 3D neon-style effect
• + High resolution with clean, sharp food photography aesthetics
• + Well-composed starburst sticker that fits the ad layout

• − The main burger is mostly assembled rather than fully 'exploded' as requested
• − Some floating sauce blobs look slightly artificial

Vidu Q2

• + Better 'exploded' effect with dramatic spacing between all layers
• + Fiery, glowing text effect matches the prompt requirements perfectly
• + Dynamic sense of motion with sauce splashes and embers

• − The currency symbol is incorrect (rendered as an 'E' or '#' cross instead of '€')
• − The bottom bun texture looks slightly less photorealistic than the top

Qwen Image

• + Excellent text legibility and spelling accuracy for almost all required items.
• + Clean and realistic chalkboard aesthetic with smudges and natural wood frame.
• + Good layout and spacing between menu items.

• − Year is incorrectly rendered as 20026 instead of 2026.
• − Text looks slightly more like a digital font than authentic chalk handwriting.

Vidu Q2

• + Authentic chalk texture with varied stroke thickness and realistic dustiness.
• + Followed the year prompt correctly (2026).
• + Captures the 'handwritten' request more effectively with natural slants and variations.

• − Significant spelling errors and 'AI gibberish' in the menu items (e.g., 'Truffe Musshoom', 'Octopd wpiln').
• − Inconsistent pricing characters and messy footer text.

Qwen Image

• + Clean, cinematic composition with realistic lighting for the astronaut's suit.
• + High resolution with sharp details on the planetary background.

• − Failed the specific prompt constraint of having the horse on top of the astronaut.
• − Anatomical issues with the horse's legs, specifically the rear leg merging into the body.

Vidu Q2

• + Beautiful surreal aesthetic with a cosmic, ethereal horse design.
• + Rich colors and intricate details in the nebulae and starry background.

• − Failed the specific prompt constraint of having the horse on top of the astronaut.
• − Minor artifacting on the reins where they meet the astronaut's hands.

Qwen Image

• + Excellent photorealistic texture on the capybara's fur
• + Accurate depiction of a yellow taxi driver cap and dark jacket
• + Cleaner facial rendering for the human passenger

• − The paws on the steering wheel look more like primate hands than capybara paws
• − The passenger is visually very close to the driver due to the framing

Vidu Q2

• + Better composition showing more of the car interior and the street environment
• + Paws on the steering wheel look more anatomically appropriate for a rodent
• + Captures a wider range of cinematic lighting from the city

• − The passenger's face is slightly distorted and less detailed
• − The capybara's head looks somewhat pasted onto the body
• − Reflection artifacts on the bottom left corner

Qwen Image

• + Perfect text rendering of the event details at the bottom
• + Excellent layout with a spooky, cinematic atmosphere
• + Successfully includes every requested element including the thorn border, scroll, and twisted trees

• − One minor typo in the large gothic title text ('Halle Party Invitation')
• − The bats are a bit simplistic in design

Vidu Q2

• + Strong gothic aesthetic for the borders and twisted trees
• + Vibrant jack-o-lantern rendering with good lighting effects

• − Significant spelling errors in both titles and scroll text ('Intovztion', 'invieed', 'fiigts')
• − Incorrect event details, displaying the wrong year and time formatting ('30.70.2025', 'Tmm')
• − The composition feels slightly more cluttered compared to the other model

Qwen Image

• + Excellent typography with clean, bold text rendering.
• + Highly effective isometric perspective and diorama composition.
• + Soft, pleasing 3D cartoon textures that perfectly match the requested aesthetic.

• − The 3D flag on the diorama base is a bit large compared to the requested 'small icon' next to text.

Vidu Q2

• + Realistic PBR-style materiality on the seafood and plate surfaces.
• + Good adherence to the request for dark bold text and central alignment.
• + Detailed modeling of disparate sushi types.

• − Composition feels slightly cluttered compared to the 'minimal' request.
• − The diorama base has odd artifacts/lumps on the corners.
• − The flag icon is awkwardly attached to the letter 'N' like a flagpole.

Qwen Image

• + Excellent soft fur texture and realistic animal renderings.
• + Clean composition with clear focus on all four requested animals.
• + Beautiful lighting with coherent god rays and dew effects.

• − Static posing that feels more like a portrait than a 'tumbling' action scene.

Vidu Q2

• + Successfully captures the 'tumbling' and 'playfully chasing' action described in the prompt.
• + Lush, expansive meadow environment with a high density of flowers.

• − Anatomical issues including a dog with an extra-long body and a bunny with a cat-like tail.
• − Inconsistent animal counts, showing two puppies instead of one.
• − Visual artifacts and clipping in the grass and butterflies.

Qwen Image

• + Excellent adherence to color scheme
• + Clear cloche illustration with simple steam
• + Minimalist vector style

• − Overlapping text makes portions unreadable

Vidu Q2

• + Elegant banner design
• + Nice subtle texture on background

• − Significant text spelling errors
• − Multiple redundant text elements
• − Cloche handle is malformed

Qwen Image

• + Stronger adherence to the navy-heavy NASA color palette.
• + Good use of flat vector graphics and icon silhouettes.
• + Text rendering is surprisingly legible for names like Armstrong and Aldrin.

• − Included instructions in the text output (Stop at landing).
• − Steps are jumbled and disorganized numerically (e.g., step 2 is next to step 1).
• − Nonsensical text integration for labels like 'Sarth Orbit Wicon)'.

Vidu Q2

• + Layout is very clean and structured as a proper infographic grid.
• + Visual style is consistent with high-quality vector illustrations.
• + The sequence of icons logically progresses from top-left to bottom-right.

• − Text consists almost entirely of gibberish strings.
• − Light gray palette lacks the bold contrast requested by the 'navy' color prompt.
• − Included 5 astronauts instead of the 3 specified in the supporting details.