FLUX.1 Kontext [pro] Black Forest Labs vs GPT Image 1.5 OpenAI

FLUX.1 Kontext [pro]

• + Excellent photographic realism with a shallow depth of field.
• + Perfectly captures the 'soft window light' lighting request.
• + Clean, modern aesthetic with realistic materials.

• − The sphere is floating unnaturally in the center and has a fuzzy, felt-like texture rather than appearing as a standard sphere.
• − The glass cube looks more like a frame than a solid glass object.

GPT Image 1.5

• + Highly accurate adherence to all spatial instructions.
• + The blue sphere features realistic light reflections and a ground reflection.
• + Excellent rendering of light refraction through the glass cube.

• − The plant is slightly more generic in shape compared to Model A.
• − The tabletop texture is a bit more repetitive than Model A.

FLUX.1 Kontext [pro]

• + Excellent natural skin textures and realistic facial features.
• + Captures a very cinematic color grade and atmosphere.
• + Includes the requested motion blur on background vehicles.

• − The subject appears to be riding or holding the bike rather than actively 'repairing' it.
• − The bicycle handlebar and cable anatomy is somewhat mangled and unrealistic.

GPT Image 1.5

• + Perfectly adheres to the 'repairing' action with tools and a crouching pose.
• + Excellent rendering of wet pavement reflections and rain on the bicycle frame.
• + Composition feels very intentional and high-quality for a street photo.

• − The background car is sharp and lacks the requested motion blur.
• − The subject's face is partially obscured by the hat and pose, though it fits the scene.

FLUX.1 Kontext [pro]

• + Excellent high-definition skin texture and realistic eye detail
• + Very subtle and natural integration of warm torchlight on the armor
• + Clean, professional photographic composition with a balanced depth of field

• − Missed the 'hair braided with small beads' requirement, showing only one clasp
• − The character looks relatively clean compared to the 'battle-worn' and 'dirt on skin' prompt requirements

GPT Image 1.5

• + Strong adherence to all prompt details including multiple beads in braids and visible scars/dirt
• + Exceptional textures on the leather straps and the cloth underlayer
• + Distinct 'battle-worn' aesthetic that feels more authentic to the character description

• − The lighting is a bit harsh, causing some overexposure on the metallic highlights
• − The hair strands near the top left appear slightly noisy/frizzy

FLUX.1 Kontext [pro]

• + Features very bold and aesthetically pleasing modern typography
• + Adheres well to the minimalist white background and clean professional layout
• + Food photography is high quality and nicely integrated into the design

• − The text is nonsensical gibberish
• − Food items do not always match the section titles (e.g., pizza under 'Mains')
• − Pricing numbers are inconsistent and confusing

GPT Image 1.5

• + Text is perfectly legible and realistic for a restaurant
• + Food photos are accurately categorized into the correct sections
• + Organized grid of images provides a rich visual experience

• − The layout is more traditional and less 'minimalist' than requested
• − Design borders on generic stock imagery

FLUX.1 Kontext [pro]

• + Excellent photographic rendering of the burger texture and dripping cheese.
• + Clean, modern typography that is easy to read.
• + Vibrant color palette and embers create a high quality 'food ad' look.

• − Failed to include the starburst for the price element.
• − The burger is not really 'exploded', just slightly separated.
• − Repeated the price text twice at the bottom which was not requested.

GPT Image 1.5

• + Successfully captured the 'exploded' view with widely suspended components.
• + Included the requested starburst for the price tag.
• + Captured all prompt elements including the specific layout of text.

• − The 'LIMITED TIME ONLY' text is cut off at the bottom of the frame.
• − Visual quality has a slightly over-sharpened, digital appearance compared to Model A.
• − Some elements, like the flying tomato chunks, look a bit cluttered.

FLUX.1 Kontext [pro]

• + Excellent text legibility and spelling accuracy
• + Realistic chalk texture and board frame
• + Good adherence to the multi-line list format requested

• − The title is not in 'elegant cursive' as requested, but rather a blocky font
• − Includes a spelling error 'foiur' at the bottom

GPT Image 1.5

• + Accurately follows the 'elegant cursive' request for the title
• + Highly realistic chalk smudge and grain texture
• + Completed the truncated menu item description naturally

• − Slightly less crisp text detail compared to the other model
• − Composition is a bit bottom-heavy with more empty space in the middle

FLUX.1 Kontext [pro]

• + Perfectly follows the specific role-reversal prompt of horse-on-top.
• + Creative surrealism with an astronaut as the mount.
• + Cinematic lighting and high-quality textures on the space suit.

• − Internal logic error where the astronaut mount has horse hooves for feet.
• − Small miniature astronaut rider on the horse's back is slightly confusing.

GPT Image 1.5

• + Excellent anatomical detail on the horse and astronaut equipment.
• + Dynamic composition with a lunar landscape and cosmic background.
• + Clear, sharp resolution across the entire image.

• − Failed the core prompt instruction of 'horse on top, not vice versa'.
• − Standard trope of astronaut on horse without the requested surreal reversal.

FLUX.1 Kontext [pro]

• + Excellent fur texture and lighting on the capybara.
• + Sharp focus on the subject with realistic car door framing.

• − The passenger is on a phone call rather than looking down at her phone as requested.
• − The perspective hides one of the front paws, making it unclear if both are on the wheel.

GPT Image 1.5

• + Perfect adherence to all prompt details including the phone usage and bored expression.
• + Strong composition showing both paws on the steering wheel and a clear view of the passenger.
• + Highly detailed taxi driver cap that reflects a classic NY aesthetic.

• − Slightly lower contrast making the background feel a bit more washed out than the foreground.

FLUX.1 Kontext [pro]

• + Excellent text readability and sharpness
• + Intricate spiderweb border design
• + Strong contrast with the glowing jack-o-lantern

• − Includes some hallucinated text lines in the event details
• − The background is very dark, losing some 'twisted tree' detail

GPT Image 1.5

• + Perfect text accuracy with no hallucinations
• + Rich vintage texture and cinematic lighting
• + More visible 'thorns' in the border as requested

• − Text is slightly less crisp than Model A
• − Jack-o-lantern light has a minor digital noise pattern

FLUX.1 Kontext [pro]

• + Excellent full and thick hair density
• + Maintains original lighting and jacket details almost perfectly

• − Changes the facial features significantly, making the person look like a different individual
• − Glasses frames changed color from light brown to black

GPT Image 1.5

• + Preserves the original facial features much more accurately than the competitor
• + Provides a very natural-looking curly texture and realistic hairline
• + Maintains the original glasses and clothing details perfectly

• − The transition from the sideburns to the new hair is slightly less seamless than Model A

FLUX.1 Kontext [pro]

• + Excellent adherence to the 'minimal' requirement
• + Clean and playful 3D cartoon style
• + Perfect alignment of the requested text and flag icon

• − The salmon texture looks slightly more like plastic than realistic PBR
• − The rice grains feel a bit simplified and rounded

GPT Image 1.5

• + Impressive realistic PBR textures on the teapot and sushi
• + Rich detail in the diorama base
• + Excellent text rendering and flag icon accuracy

• − Ignored the 'minimal' instruction by adding many extra props like a teapot and soy sauce
• − The composition is much more crowded than requested

FLUX.1 Kontext [pro]

• + Excellent preservation of the source subject's identity and original clothing.
• + Applies a clean cartoon/caricature style while maintaining the selfie composition.

• − Completely failed to include the requested profession (TV anchor), dogs, or hockey themes.
• − The edit is more of a stylization than a thematic caricature.

GPT Image 1.5

• + Successfully incorporates all requested elements: TV anchor desk, microphone, news graphics, dogs, and hockey imagery.
• + Effectively uses caricature proportions by exaggerating the head and smile while maintaining a high level of detail.
• + Creates a humorous and busy scene that fits the 'exaggerated' prompt.

• − Less preservation of the source image's specific clothing and pose.
• − The face, while recognizable, leans more toward a generic caricature than a direct translation of the source person.

FLUX.1 Kontext [pro]

• + Excellent soft lighting and atmosphere.
• + High level of detail in the fur textures.
• + Very aesthetically pleasing composition with a consistent style.

• − Failed to generate a distinct bunny; the third animal appears more like a kitten-rabbit hybrid.
• − The animals are sitting rather than tumbling or playfully chasing as requested.

GPT Image 1.5

• + Successfully captured the 'tumbling' and 'playfully chasing' action described in the prompt.
• + Identifiable representation of all four requested animals (dog, cat, bunny, fox).
• + Beautiful god rays and dew sparkle effects that match the 8K masterpiece requirement.

• − The fox's paw in the bottom right corner is somewhat distorted and dark.
• − Slightly less 'photorealistic' and more illustrative/saturated than Model A.

FLUX.1 Kontext [pro]

• + Excellent preservation of the source image's composition and poses.
• + Clearly captures the late 80s/early 90s Ghibli aesthetic with clean line work and watercolor textures.
• + Maintains the distinct blue plaid pattern of the man's shirt very accurately.

• − The woman in the foreground's facial expression is slightly more generic than the source's look.
• − The paper texture is a bit uniform across the entire canvas.

GPT Image 1.5

• + Successfully captures the dreamy, soft-glow lighting often found in modern Ghibli or Makoto Shinkai films.
• + Good use of pastel colors and a nostalgic, warm color grade.
• + Preserves the character's positions and essential outfits well.

• − The man's hand disappears into his back/pants unnaturally.
• − The blurring effect on the foreground woman is a bit excessive compared to the source and the style.
• − Loss of some fine detail in the background due to the heavy lighting bloom.

FLUX.1 Kontext [pro]

• + Excellent source preservation, maintaining the identity and features of the woman and dog perfectly.
• + Subtle, realistic blowing hair that integrates seamlessly with the original style.

• − The 'flying leaves' instruction is barely met, with only a few small green dots visible.
• − Overall motion is less 'dynamic' than requested.

GPT Image 1.5

• + Strong adherence to the 'flying leaves' prompt with numerous leaves of varied colors.
• + High degree of energetic motion in the hair.

• − Significant loss of source preservation; the woman's face has been noticeably altered.
• − The leaf placement looks like a flat overlay rather than being integrated into the 3D space of the scene.

FLUX.1 Kontext [pro]

• + Excellent typography with a professional serif font and accurate accented characters.
• + Perfectly captures the 'subtle texture on light background' requested in the prompt.
• + Clean, minimalist vector aesthetic suitable for a real restaurant logo.

• − Includes a minor spelling error in the banner ('EEST' instead of 'EST').
• − Steam interpretation is very basic compared to the rest of the illustration.

GPT Image 1.5

• + Beautifully rendered cloche with sophisticated lighting and shading.
• + Perfect spelling on all text elements including the banner.
• + Dynamic and elegant steam illustration.

• − Ignored the request for a 'light background', opting for solid black.
• − The layout feels a bit cramped compared to the balanced spacing of the first image.
• − Typography for 'Caffè' is slightly less clean than the main brand name.

FLUX.1 Kontext [pro]

• + Features a bold, high-quality main title
• + Includes clean, modern vector styling for the lunar terrain

• − Confused astronomical objects, placing Saturn's rings on the rocket and the Moon
• − Text labeling is disorganized and partially nonsensical (e.g., '2, far Collins')
• − Failed to follow the requested sequential 6-step infographic structure

GPT Image 1.5

• + Strictly adheres to all 6 requested steps in a logical infographic layout
• + Text rendering is clear, accurate, and correctly positioned for every stage
• + Maintains a consistent, clean flat-vector icon style across all panels

• − The 'Tranquillity' speech bubble is slightly redundant with the landing label
• − The inclusion of astronaut silhouettes was not explicitly requested, though it fits the theme