FLUX.1 Kontext [max] Black Forest Labs vs GPT Image 2 OpenAI

FLUX.1 Kontext [max]

• + Excellent handling of light and caustic reflections on the wooden table.
• + Highly realistic glass transparency and thickness.
• + The plant is clearly visible through the glass as requested.

• − The sphere has a textured, glitter-like surface rather than a smooth finish.
• − The text on the book spine is nonsensical.

GPT Image 2

• + Perfect adherence to all spatial instructions and object descriptions.
• + The sphere has a clean, matte blue finish that feels very cohesive.
• + Excellent composition with the window clearly visible to establish the light source.

• − The glass cube edges appear slightly thinner and less physically substantial than Model A.
• − Slightly less complex light interaction on the table surface.

FLUX.1 Kontext [max]

• + Excellent depiction of rain with realistic streaks and wet surface reflections.
• + Dynamic composition with 'imperfect' cinematic framing requested in the prompt.
• + Superior textures on the man's hands and face, feeling very natural.

• − The man's ethnicity appears somewhat ambiguous rather than specifically Japanese.
• − The rain effects are a bit heavy, slightly obscuring the background motion blur.

GPT Image 2

• + Successfully captures the specific Japanese setting including legible kanji/signage.
• + Good motion blur on the passing white car in the background.
• + Includes additional realistic elements like a tool kit and a worn, weathered bicycle.

• − The 'light rain' is barely visible, lacking the atmospheric quality of Image A.
• − The pavement doesn't look particularly wet or reflective.
• − The lighting is a bit flat and less cinematic than requested.

FLUX.1 Kontext [max]

• + Excellent depiction of warm torchlight reflecting off the armor and skin.
• + Highly detailed engraving on the plate armor with convincing metal textures.
• + Strong adherence to the 'battle-worn' descriptor with visible sweat, grime, and intense eyes.

• − The beads in the hair are missing or very indistinct.
• − Skin texture on the forehead appears slightly oversaturated or unnaturally shiny.

GPT Image 2

• + Perfect adherence to the 'hair braided with small beads' requirement.
• + Exceptional facial detail with realistic freckles, faint scars, and dirt.
• + Clearly visible textured leather straps and cloth underlayers as requested.

• − The lighting is a bit more muted than the 'warm torchlight' requested.
• − The depth of field is less shallow compared to Model A.

FLUX.1 Kontext [max]

• + Clean layout with effective use of white space.
• + Maintains a consistent aesthetic for a pizza-focused casual dining environment.

• − Text is largely illegible gibberish.
• − The grid is repetitive, showing almost exclusively pizza even for different sections.
• − The font choice for section headers is a script font rather than bold sans-serif.

GPT Image 2

• + Perfect adherence to section requirements (Appetizers, Pizza, Mains) with distinct food photos for each.
• + Exceptional text rendering with coherent item names, descriptions, and prices.
• + Follows all stylistic cues including bold sans-serif fonts and vibrant accents.

• − The layout is slightly more dense than Model A, though still professional.

FLUX.1 Kontext [max]

• + Excellent text legibility and clean font rendering.
• + Good photorealistic texture on the central burger patty and bun.

• − Fails to create a truly 'exploded' view, with most ingredients still stacked.
• − Missing the starburst element for the price tag.

GPT Image 2

• + Perfect adherence to the 'exploded' view requirement with vertical separation.
• + Dynamic fiery text effects and starburst element exactly as requested.
• + Thorough integration of all components including sauces and glowing embers.

• − Slightly cluttered composition with the text overlapping some sparks.
• − The price text is a bit thick, making the decimal point small.

FLUX.1 Kontext [max]

• + Perfectly legible and accurate text rendering for the entire prompt.
• + Good chalk smudging effects on the background for realism.
• + Clear, high-contrast composition.

• − The text looks a bit too clean, like a digital chalk-style font rather than organic handwriting.
• − Failed to follow the instruction for 'elegant cursive' for the title.

GPT Image 2

• + Excellent chalk texture and realistic varied handwriting style.
• + Successfully followed the request for elegant cursive in the header.
• + Warm, cozy café lighting enhances the atmospheric quality.

• − The text is slightly less legible due to the thinness of the chalk strokes.
• − One small typo in the word 'Special' (looks like 'Specials' with a slightly merged 's').

FLUX.1 Kontext [max]

• + Excellent anatomical rendering for both the horse and astronaut
• + High-quality cinematic lighting and background composition
• + Clear, high-resolution textures

• − Failed to follow the logical reversal requested in the prompt ("horse on top")

GPT Image 2

• + Perfectly adhered to the complex prompt instruction of having the horse on top of the astronaut
• + Highly surreal and creative interpretation of the concept
• + Detailed textures on the spacesuit and lunar surface

• − Mismatched scale between the horse's upper body and its tiny hooves
• − Awkward anatomical merger where the horse's legs meet the astronaut's shoulders

FLUX.1 Kontext [max]

• + High resolution and realistic fur texture on the capybara.
• + The yellow taxi cab lighting on the vehicle frame is very convincing.

• − The woman in the back is holding a phone to her ear like a call, rather than looking at it as requested.
• − The capybara's paws are not clearly placed on the steering wheel as requested.

GPT Image 2

• + Perfect adherence to the pose, showing both paws on the steering wheel.
• + Excellent depiction of the passenger looking at her phone with a bored expression.
• + The interior composition feels more like a real taxi ride with a clearer view of the back seat.

• − The capybara's fur looks slightly less defined compared to Model A.
• − The proportions of the capybara's head to the steering wheel are a bit oversized.

FLUX.1 Kontext [max]

• + Excellent text legibility and size hierarchy
• + Very clean border design that incorporates webs and thorns effectively
• + Perfect adherence to the specified date and color contrast

• − The 'Arches' location is repeated twice in the text block
• − The background is slightly less detailed than the competitor

GPT Image 2

• + Beautifully detailed gothic illustration including a castle, bridge, and full moon
• + The scroll banner is more stylistic and aesthetically pleasing
• + Text rendering is highly accurate with no repetitions

• − The event details at the bottom are quite small and harder to read at a glance
• − The border, while intricate, feels a bit cluttered compared to the main title

FLUX.1 Kontext [max]

• + Excellent adherence to the 'minimal garnish' request
• + Smooth, refined 3D textures that match the 'soft cartoon' aesthetic well
• + Clean text rendering for 'JAPAN' and 'SUSHI'

• − Failed to include the requested flag icon
• − The background is slightly more saturated than 'light blue'

GPT Image 2

• + Perfect adherence to all prompt elements, including the flag icon
• + Higher detail in materials and textures while maintaining the diorama look
• + Very clean, bold typography that pops against the background

• − Ignores the 'minimal garnish' instruction by adding many extra elements
• − The lighting is slightly more artificial and harsh compared to Model A

FLUX.1 Kontext [max]

• + Successfully includes all four requested species with clear, distinct features.
• + Captured a soft, 'dreamy' lighting style that enhances the wholesome vibe.
• + Excellent fur textures and expressive eye glints.

• − The composition is quite static, with animals sitting rather than 'playfully chasing' or 'tumbling'.
• − The butterflies appear somewhat flat and glowy rather than realistic.

GPT Image 2

• + Excellent depiction of movement, with animals leaning forward and paws raised to simulate chasing.
• + The lighting and god rays are very dramatic and well-integrated into the scene.
• + High level of detail in the grass and wildflowers providing a sense of depth.

• − The fox kit has a slightly unusual anatomy on its front right paw.
• − The kitten's tail is positioned oddly, almost appearing to sprout from its upper back.

FLUX.1 Kontext [max]

• + Excellent adherence to the 'minimalist' keyword with clean forms
• + Perfect text rendering of the brand name and establishment year
• + Subtle, effective texture that enhances the vintage feel without clutter

• − The steam element is a bit thick and stylized compared to the cloche
• − Slightly less 'grand' than the historic nature of the real restaurant

GPT Image 2

• + Beautiful, intricate linework on the cloche and frame
• + Sophisticated typography that feels very high-end and classical
• + Warm color palette and parchment-style background are executed perfectly

• − Fails the 'minimalist' requirement by including heavy framing and flourishes
• − The font weights between the two names are slightly unbalanced

FLUX.1 Kontext [max]

• + Clean minimalist color palette and background.
• + Includes clear silhouette icons for the three astronauts.
• + The overall flat-vector style is consistent throughout the image.

• − Misses several steps requested in the prompt (only labels 4 areas).
• − Displays generic or nonsensical icons instead of the specific Saturn V and Lunar Module requested.
• − Layout is disorganized and text labels are poorly placed.

GPT Image 2

• + Exemplary adherence to the step-by-step instructions and iconography.
• + High-quality vector illustrations of the Saturn V and Lunar Module.
• + Superior layout organization with clear headings and consistent graphic design.

• − The eagle mascot in the top right corner is slightly busy compared to a strictly flat-vector style.
• − Minor lighting effects on the lunar surface deviate slightly from the 'subtle gradients only' constraint.