GPT Image 1.5 OpenAI vs Qwen Image Alibaba

GPT Image 1.5

• + Excellent adherence to lighting instructions with a clear left-side window source
• + Realistic glass thickness and internal reflections
• + High-quality textures on the red book and wooden table

• − The sphere appears slightly large relative to the prompt 'small'

Qwen Image

• + Clean, minimalist composition
• + Correct placement of all objects according to the prompt

• − The glass cube appears to have a mirrored base which wasn't requested
• − The green plant is less visible through the glass compared to Model A
• − The sphere appears to be floating rather than resting on the surface

GPT Image 1.5

• + Excellent adherence to the 'imperfect framing' and '50mm lens' look
• + Highly realistic skin textures and wet surfaces
• + Includes specific details like a tool box and realistic bicycle mechanics

• − The motion blur on the car is slightly static compared to the rain effect

Qwen Image

• + Good inclusion of background motion blur on the white car
• + Correctly captures the 'red bicycle' and 'elderly Japanese man'

• − The bicycle rendering is physically incoherent with a floating pedal and missing chain
• − The overall image quality feels somewhat AI-smoothed and less like a professional 50mm shot

GPT Image 1.5

• + Exceptional photographic realism with lifelike skin texture and eyes.
• + Beautiful integration of warm torchlight and glowing bokeh embers.
• + Intricate detail on the engraved armor, worn leather straps, and textured cloth.

• − The hair beads are metal rings rather than the traditionally expected decorative beads.

Qwen Image

• + Accurately depicts specific colorful beads in the hair braids.
• + Clear representation of a torch as the light source.
• + Strong engraving details on the shoulder pauldron.

• − The facial scars look painted on rather than realistic skin texture.
• − The bokeh sparks appear as artificial geometric star shapes.
• − Lower overall visual fidelity compared to the rival image.

GPT Image 1.5

• + Excellent text rendering with no spelling errors or hallucinations.
• + Photorealistic food imagery that matches the specific menu items listed.
• + Logical and professional hierarchy with clear pricing and descriptions.

• − The layout is a bit crowded with narrow margins.
• − The 'vibrant accents' are limited to simple colored underlines.

Qwen Image

• + Strong minimalist aesthetic with a clear grid system for photos.
• + Creative use of color blocking behind some grid items.
• + Bold typography fits the modern sans-serif requirement.

• − Illegible, nonsensical text throughout the menu.
• − Food images are repetitive and less realistic than the competitor.
• − Failed to create three distinct sections for appetizers, pizza, and mains as requested.

GPT Image 1.5

• + Excellent adherence to the 'exploded' request with detailed vertical separation of ingredients.
• + Superior photorealistic textures on the meat patty, bun, and vegetables.
• + Integrated the fiery, glowing effect into the typography and starburst perfectly as requested.

• − The overall composition is a bit crowded with many small floating particles.
• − The 'MAGIC BURGER' text is slightly clipped at the top edges.

Qwen Image

• + Clean, professional graphic design layout with good breathing room.
• + Highly legible secondary text ('LIMITED TIME ONLY').
• + Effective use of depth of field in the background flames.

• − Failed the architectural part of the prompt; the burger is mostly assembled rather than 'exploded'.
• − The glowing effect on the text feels more like neon tubing than the requested fiery effect.
• − Less photorealistic detail in the food textures compared to the competitor.

GPT Image 1.5

• + Excellent text rendering with perfect spelling for all complex menu items.
• + Authentic chalk texture with realistic dusty smudges and varying stroke opacity.
• + Completed the third menu item correctly despite the truncated prompt.

• − The composition is a very tight close-up, lacking the 'cozy café' environmental context.

Qwen Image

• + Good inclusion of the café environment in the background.
• + Clean, legible handwriting with a pleasant slant.

• − Incorrect year rendered as '20026' instead of '2026'.
• − Spelling errors in menu items such as 'Risoto' instead of 'Risotto'.
• − Text appears more like a digital brush than realistic chalk on board.

GPT Image 1.5

• + Excellent surface textures with high-frequency detail in the space suit and horse's fur.
• + Dynamic composition with a cinematic sense of motion and scale.
• + Rich, complex background featuring celestial bodies, nebulae, and lunar hardware.

• − Completely failed the negative constraint to have the horse on top of the astronaut.

Qwen Image

• + Clean, clear lighting and a minimalist, surreal aesthetic.
• + Realistic anatomy for the horse and astronaut gear.
• + Smooth rendering with minimal digital artifacts.

• − Completely failed the negative constraint to have the horse on top of the astronaut.
• − Less detailed background compared to the competitor.

GPT Image 1.5

• + Excellent photorealistic fur texture and lighting integration.
• + Correct taxi hat and steering wheel placement for a capybara.
• + The businesswoman's bored expression perfectly captures the requested irony.

• − The paws on the steering wheel look slightly like human fingers in gloves rather than natural capybara feet.

Qwen Image

• + Clean, modern composition with a clear view of the city background.
• + Accurate jacket and name tag detail on the driver.

• − The paws are anatomically incorrect, looking more like primate hands.
• − Text rendering on the roof sign says 'YOXI' instead of 'TAXI'.
• − The capybara's head is not looking toward the road, creating an unnatural driving posture.

GPT Image 1.5

• + Perfect text rendering for all requested strings including long sentences and dates.
• + Highly detailed and consistent gothic aesthetic with a beautiful thorn and web border.
• + Excellent cinematic lighting and composition that feels like a professional poster.

• − The parchment texture is very dark, which may slightly impact readability of the bottom text.

Qwen Image

• + Successfully incorporates all requested elements like twisted trees, bats, and thorns.
• + Clean layout with clear separation between the illustration and the document edges.

• − Significant typos in the main title, rendering it as 'Halloforty Invitation'.
• − The thorn border is a flat graphic rather than an integrated part of the cinematic scene.
• − Visual style is slightly more cartoonish and less 'vintage gothic' than requested.

GPT Image 1.5

• + Excellent PBR material rendering with realistic wood, ceramic, and food textures.
• + Perfect text rendering and alignment for 'JAPAN' and 'SUSHI'.
• + High level of detail in the miniature diorama elements like the teapot and soy sauce bottle.

• − The scene is a bit more realistic than the requested '3D cartoon' style.
• − The diorama base is quite busy compared to the 'minimal' request.

Qwen Image

• + Perfectly captures the '3D cartoon' and 'soft refined texture' aesthetic requested.
• + Better adherence to the 'minimal garnish and plate' instruction.
• + Clean isometric composition with a clear miniature toy-like feel.

• − Text rendering is slightly less crisp and has a strange orange artifact under the flag icon.
• − Lower level of intricate detail compared to Model A.

GPT Image 1.5

• + Excellent fur texture and fine detail rendering on all four animals.
• + Superior lighting and atmosphere with convincing god rays and dew sparkles.
• + Dynamic and joyful composition that perfectly captures the 'tumbling together' prompt.

• − The kitten has an anatomical anomaly with an extra-looking limb or paw placement under its chin.
• − The fox kit's face looks slightly more like a dog than a wild fox.

Qwen Image

• + Clean composition with clearly defined subjects and a playful arrangement.
• + Accurate representation of the specific species requested with distinct features.
• + Good use of bokeh and sparkles in the meadow to create a magical atmosphere.

• − The golden retriever puppy appears much flatter and less detailed in its fur compared to the other animals.
• − The kitten's front right leg is anatomically distorted and looks disconnected from the body.
• − Lighting is a bit generic and lacks the 'hyper-photorealistic' depth requested.

GPT Image 1.5

• + Excellent typography with perfect spelling and professional layout.
• + High-quality vector aesthetic with sophisticated shading and texture.
• + Correctly follows the vintage cloche and banner concept.

Qwen Image

• + Applies a nice subtle paper texture to the background.
• + Good alignment with the requested warm brown and cream color scheme.

• − Serious spelling and typography errors, overlapping text characters.
• − Loss of 'minimalist' feel due to messy letter placement.
• − The cloche illustration is very basic compared to its counterpart.

GPT Image 1.5

• + Excellent text rendering with no spelling errors.
• + Clear, logical panel-based layout following all six requested steps.
• + Consistently applied flat-vector style with high-quality icons.

• − The rocket in the first panel is missing its top nose cone section.
• − The horizon in the Launch panel looks more like Mars than a NASA-inspired Earth red.

Qwen Image

• + Successfully captured the requested NASA-inspired color palette.
• + Good use of negative space in a single-page infographic style.

• − Numerous spelling errors in text (e.g., 'Sarth Orbit', 'Apoll', 'Aldlin').
• − Failed to include all 6 distinct steps, skipping some and mislabeling others.
• − Included prompt instructions like '(Stop at landing)' as actual text on the image.