GPT Image 1 OpenAI vs Stable Diffusion 3.5 Medium Stability AI

GPT Image 1

• + Excellent adherence to the complex spatial prompt.
• + High visual quality with realistic textures on the book, glass, and wood.
• + Natural and soft window lighting appearing from the requested direction.

• − The plant is more besides the cube than behind it, though it remains visible through the pane.

Stable Diffusion 3.5 Medium

• + Successfully places the plant behind the object as requested.
• + Clean rendering of the glass material.

• − Completely failed the spatial instruction for the red book, placing it inside the cube instead of on top.
• − The blue sphere is sitting on the book rather than just being inside the cube.
• − The composition is cropped poorly at the top.

GPT Image 1

• + Excellent skin texture and hyper-realistic facial details
• + Accurate shallow depth of field with 50mm look
• + Consistent lighting and rain effect on surfaces

• − Anatomical issue with the hand merging into the bicycle chain and spokes
• − Missing motion blur from passing cars requested in prompt
• − Car headlights in the background are out of focus but static rather than blurred

Stable Diffusion 3.5 Medium

• + Stronger composition for a 'candid' street photo
• + Better implementation of the 'motion blur from passing cars' request
• + Captures the atmosphere of a rainy Japanese street very well

• − Poor facial detail and skin texture compared to Model A
• − Artificial-looking artifacts around the man's hands and the bicycle basket
• − Low clarity and a somewhat 'smeared' look rather than intentional shallow focus

GPT Image 1

• + Excellent depiction of battle-worn textures including skin grime and aged metal.
• + Atmospheric lighting with consistent warm torchlight reflections.
• + Superior focus and shallow depth of field effects.

• − Lacks visible leather straps mentioned in the prompt.
• − The 'beads' in the hair are somewhat subtle and blend into the braid.

Stable Diffusion 3.5 Medium

• + Clear representation of leather straps and cloth underlayers.
• + Included distinct hair beads and more complex braiding patterns.
• + Sharp facial features and lifelike eye color.

• − The lighting feels more artificial and less like ambient torchlight.
• − The armor engraving and skin texture look too clean for a 'battle-worn' character.
• − Bokeh sparks appear as floating circles that don't integrate naturally with the scene.

GPT Image 1

• + Excellent typography rendering with clean, readable sans-serif fonts
• + High-quality, appetizing food photography with vibrant colors
• + Logical, minimalist layout that adheres perfectly to the grid request

• − Minor spelling errors in descriptive text like 'Apperoiation descrigion'
• − Limited number of menu items displayed compared to a full page

Stable Diffusion 3.5 Medium

• + Successfully creates a full-page menu concept
• + Good use of space and multiple food photos
• + Captures the professional casual dining aesthetic

• − Text is completely illegible and visually garbled
• − Food images lack fine detail and clarity
• − Poor source preservation on fonts, looking more like symbols than actual characters

GPT Image 1

• + Excellent adherence to the 'exploded' concept with all components clearly separated and suspended.
• + High-quality text rendering with the requested glowing, fiery effect.
• + Photorealistic textures on the food items, particularly the bun and the lettuce.

• − The price text is incorrect, displaying '.99' instead of '€6.99'.
• − The starburst around the price is a bit simplified compared to the rest of the professional-grade typography.

Stable Diffusion 3.5 Medium

• + Accurate text spelling including the price and the currency symbol.
• + Strong background dynamics with literal flames and flying debris.
• + Clean, modern graphic design for the price starburst.

• − Failed the core 'exploded burger' requirement; the burger is mostly assembled rather than suspended in components.
• − The text does not have the requested 'fiery, glowing effect' and looks like flat overlays.
• − Visible artifacts and blurring where the meat meets the cheese and lettuce.

GPT Image 1

• + Excellent text legibility and accuracy
• + Convincing chalk texture on the letters
• + Perfect adherence to the specific menu items and prices requested

• − Font looks slightly too uniform for 'natural variations'
• − Title is not in cursive as requested

Stable Diffusion 3.5 Medium

• + Beautiful artistic chalk style and layout
• + Included a wooden frame that enhances the cozy café aesthetic
• + Better variation in letter size and artistic flourishes

• − Severe spelling errors throughout the board
• − Failed to follow the requested menu text and prices
• − Layout is disorganized and cluttered with gibberish

GPT Image 1

• + Excellent cinematic lighting and texture on the space suit and horse hide.
• + Higher artistic coherence with a consistent 'surreal' aesthetic.
• + Better anatomical rendering of the horse compared to the competitor.

• − Prompt specified 'horse on top' (implying the horse riding the astronaut), but the astronaut is riding the horse.

Stable Diffusion 3.5 Medium

• + Clear, high-contrast composition against the planet's atmospheric glow.
• + Accurate space suit detailing and vibrant color palette.

• − Failed the negative constraint; the astronaut is riding the horse instead of the inverse.
• − Significant anatomical errors with the horse's legs, including extra segments and warped hooves.
• − The horse's neck and head join the body at an awkward, unrealistic angle.

GPT Image 1

• + Perfect adherence to the phone-using businesswoman instruction
• + High photorealistic quality with naturalistic lighting and textures
• + Accurate representation of capybara anatomy holding a steering wheel

• − The perspective makes the capybara look slightly large for the cab interior

Stable Diffusion 3.5 Medium

• + Good centered composition that highlights the main subject
• + Vibrant colors and clear details on the capybara's face

• − Failed to include the businesswoman looking at her phone
• − Capybara's paws look more like bird claws or generic paws rather than capybara feet
• − The steering wheel is missing or poorly defined

GPT Image 1

• + Flawless text rendering with correct spelling and typography.
• + Cohesive and elegant vintage gothic aesthetic.
• + All requested elements including bats, trees, and scroll are integrated naturally into the composition.

• − Lighting is a bit dark, making some of the background details subtle.

Stable Diffusion 3.5 Medium

• + Bright and vibrant colors that pop against the sky.
• + Good use of the parchment paper texture and torn edges.

• − Numerous spelling errors in every line of text.
• − The composition feels cluttered with two jack-o-lanterns instead of one central one as requested.
• − The font style is more playful than the requested elegant gothic style.

GPT Image 1

• + Excellent adherence to the 'small raised diorama base' instruction.
• + Perfect text rendering and inclusion of the flag icon as requested.
• + Very clean 3D cartoon aesthetic with soft, refined textures.

• − The salmon texture on the nigiri is slightly repetitive.

Stable Diffusion 3.5 Medium

• + Successfully renders the sushi on a plate with requested text.
• + Good color vibrancy in the food items.

• − Missed the 'small raised diorama base' and 'flag icon' instructions.
• − Text layout is poor, with 'JAPAN' being much smaller than 'SUSHI' contrary to the prompt.
• − Low visual clarity with noticeable artifacts and blurry text elements.

GPT Image 1

• + Excellent adherence to the prompt by including all four distinct animals.
• + Dynamic composition with a strong sense of movement and 'tumbling' interaction.
• + Beautiful lighting with visible god rays and realistic textures on the fur.

• − The fox's front paw has a slightly awkward anatomical structure.
• − Minor clipping where the kitten's paw overlaps the rabbit.

Stable Diffusion 3.5 Medium

• + Vibrant color palette with high-contrast floral elements.
• + Clean, sharp rendering of the animals' faces.

• − Failed to include the baby bunny, only showing three animals.
• − The kitten's ears are unusually long and pointed, looking more like a caracal or a hybrid.
• − Static composition that lacks the 'tumbling' and 'chasing' action requested in the prompt.

GPT Image 1

• + Excellent text rendering with no spelling errors
• + Follows the minimalist vector emblem style perfectly
• + High contrast and clean lines suitable for a logo

• − Ignored the 'light background' instruction using black instead
• − Minimal visual texture compared to the request

Stable Diffusion 3.5 Medium

• + Successfully used a light background with subtle texture
• + Sophisticated vintage illustration style
• + Includes the 'banner' element and 'cream' tones

• − Multiple spelling errors in the main name and date
• − The cloche is poorly defined and resembles a helmet or building
• − Overly complex for a 'minimalist' logo request

GPT Image 1

• + Excellent typography with legible names and headings
• + Clean flat-vector aesthetic that matches the requested NASA palette
• + Includes accurate supporting details like the astronaut names

• − Confusing layout order where labels don't always align with the icons below them
• − Spelling error in 'EARLLUNAR'

Stable Diffusion 3.5 Medium

• + Dynamic composition with a strong sense of trajectory and scale
• + Follows the sequential list format relatively well on the vertical axis

• − Poor typography with significant illegible 'loreum ipsum' style text
• − Failed to follow the specific iconography requests (e.g., Saturn V icon, Lunar Module on surface)
• − Messy visual elements that deviate from the requested 'crisp lines' flat-vector style