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

FLUX.1 Kontext [max]

• + Excellent handling of complex light refraction and shadows on the table surface.
• + Highly realistic textures, especially the glittery surface of the sphere and the wood grain.
• + Strong adherence to the spatial prompt regarding the plant behind the glass.

• − The book's placement looks slightly floating/unbalanced on the glass edge.

GPT Image 1 Mini

• + Clean, minimalist composition with accurate item placement.
• + Clear and simple representation of the blue sphere and red book.
• + Good lighting consistency from the left side.

• − The glass cube lacks realistic thickness and refraction characteristics compared to Model A.
• − The plant is very blurred, making the 'visible through the glass' instruction less impactful.
• − Texture on the wood and sphere is overly smooth/plastic.

FLUX.1 Kontext [max]

• + Excellent depiction of rain with atmospheric streaks.
• + Vibrant reflections on the wet pavement add to the cinematic quality.
• + Compelling composition with the bike spanning the foreground.

• − Physical interaction between the hands and the chain is messy and anatomically confusing.
• − Rain streaks sometimes appear as static white lines in front of the lens rather than 3D depth.

GPT Image 1 Mini

• + Natural skin texture and facial expression are highly realistic.
• + Better anatomical correctness with the hands and fingers.
• + Accurate shallow depth of field and soft background bokeh.

• − The 'light rain' is barely visible, missing the atmospheric quality of Image A.
• − The pavement is wet, but lacks the dynamic reflections requested in the prompt.

FLUX.1 Kontext [max]

• + Sublime metal engraving and reflection details.
• + Excellent skin texture with visible pores and realistic sweat/dirt.
• + Striking lifelike eyes that catch the warm lighting perfectly.

• − Hair braids look slightly like thick ropes rather than traditional hair braids.
• − The sparks in the background are a bit elongated and distracting.

GPT Image 1 Mini

• + Great 'battle-worn' aesthetic with more grit and blood on the face.
• + Sophisticated, subtle engraving on the plate armor.
• + Better integration of the braids into the character's hairstyle.

• − Lighting is a bit flat compared to the dynamic reflections in Image A.
• − The leather strap detail is less distinct than requested.

FLUX.1 Kontext [max]

• + Excellent professional photorealism in the food items
• + Realistic layout that mimics a real-world printed menu
• + Balanced use of brand accents and typography

• − Text is mostly gibberish despite looking like a real menu
• − Food variety is lacking as it is almost entirely pizza

GPT Image 1 Mini

• + Perfect text rendering for the requested sections
• + Distinct and varied food photos covering all categories in a clear grid
• + Clean, high-contrast minimalist aesthetic

• − The layout is overly simplistic and lacks descriptive text placeholder areas
• − The food images look slightly more like stock graphics than integrated photography

FLUX.1 Kontext [max]

• + Excellent photorealistic texture on the burger patty and buns
• + Great sense of high-speed motion with flying debris and embers
• + Clean and readable typography centered well in the layout

• − Failed to place the price in a starburst as requested
• − The burger is more of a complete sandwich with side pieces rather than a fully 'exploded' vertical stack
• − Missing the fiery/glowing effect specifically on the secondary text

GPT Image 1 Mini

• + Perfect adherence to the 'exploded' vertical stack layout
• + Captures the 'fiery, glowing' text effect much better than Model A
• + Correctly included the price within a starburst graphic

• − The burger ingredients look somewhat plasticky or artificial compared to Model A
• − The texturing on the bottom bun is slightly muddy
• − Lighting on the burger is a bit flat despite the glowing background elements

FLUX.1 Kontext [max]

• + Excellent text rendering with perfect spelling and spacing.
• + Realistic chalk texture including smudges on the board.
• + Strong environmental context with café elements visible.

• − The title is in print/block letters rather than the requested elegant cursive.

GPT Image 1 Mini

• + The chalk texture on the letters is very grainy and realistic.
• + Good layout and alignment of prices.

• − Failed to render the title in cursive as requested.
• − The handwriting looks somewhat uniform, bordering on a font style.
• − The board feels less integrated into a physical space compared to Model A.

FLUX.1 Kontext [max]

• + Excellent lighting and color contrast between the brown horse and the dark space.
• + High texture detail on the astronaut suit and horse's coat.
• + Dynamic composition with a sense of motion.

• − Failed the negative constraint; the astronaut is riding the horse instead of the horse on top.

GPT Image 1 Mini

• + Atmospheric cinematic lighting with subtle lens flare effects.
• + Clean anatomical rendering of both the astronaut and the horse.
• + Consistent starry background with good depth of field.

• − Failed the negative constraint; like Model A, it shows an astronaut riding a horse.
• − Low color saturation makes the image feel slightly flat compared to Model A.

FLUX.1 Kontext [max]

• + Excellent photorealistic texture on the capybara's fur and the leather taxi cap
• + Captures the sense of motion and vibrant New York city lights through the window
• + Superior rendering of the taxi's exterior and side-view perspective

• − The passenger is holding a phone to her ear like a call, rather than looking at it as requested
• − Only one paw is clearly making contact with the steering wheel

GPT Image 1 Mini

• + Perfect adherence to the passenger's expression and action (looking at phone)
• + Clearly shows both paws on the steering wheel as requested
• + Nostalgic, moody lighting that fits a night-time taxi theme

• − The image is overall very dark with lower contrast in the interior
• − The capybara's paw anatomy on the steering wheel looks slightly distorted

FLUX.1 Kontext [max]

• + Excellent gothic typography that feels period-appropriate.
• + Superior lighting and depth, especially the glow emanating from the jack-o-lantern.
• + Very detailed border with clear web and thorn elements.

• − Redundant text at the bottom repeats the location twice.
• − The commas in the date '30,10,2026' are a slight deviation from the requested period punctuation.

GPT Image 1 Mini

• + Perfectly follows the specific text for location and date without repetition.
• + The scroll banner design is very elegant and well-integrated into the composition.
• + Clean, readable layout that balances the vintage parchment feel well.

• − The lighting is much flatter compared to the high-contrast cinematic lighting in Model A.
• − The border details are muddier and less distinct than the thorns and webs in the competing model.

FLUX.1 Kontext [max]

• + Excellent preservation of the subject's unique facial structure and wrinkles
• + Realistic hair texture and thickness
• + High-quality skin texture and detail preservation

• − The hairline transition is slightly harsh with a faint white line artifact

GPT Image 1 Mini

• + Natural, messy hair texture that fits the desert setting
• + Good blending of the hair at the temples

• − Significant loss of source identity by smoothing skin and changing facial features
• − Alters the shape of the nose and glasses
• − Removed original details like the button and weathering on the jacket

FLUX.1 Kontext [max]

• + Excellent soft refined 3D textures on the sushi and rice.
• + Very clean typography and centered composition.
• + Natural-looking soft lighting and shadows.

• − Missed the small flag icon request.
• − The brown text color has slightly less contrast against the blue than Model B.

GPT Image 1 Mini

• + Includes all prompted elements including the small flag icon.
• + Crisp text with high contrast for readability.
• + Good adherence to the 45-degree isometric perspective.

• − The sushi looks a bit more plasticky/artificial compared to Model A.
• − The grains of rice look slightly less realistic than those in the other image.

FLUX.1 Kontext [max]

• + Successfully captures the likeness of the original subject's hair and denim shirt.
• + Incorporates all three elements: TV anchor, dog, and hockey stick.
• + Uses a clean, vibrant digital illustration style.

• − Added glasses that weren't in the original image, reducing the likeness.
• − The hockey stick is partially cut off at the top.
• − The dog looks more like a clip-art icon than a character in the scene.

GPT Image 1 Mini

• + Stronger 'caricature' style with exaggerated facial features that still mimic the original person.
• + Cleverly includes a hockey puck in addition to the stick for better thematic coverage.
• + Better integration of the dog as a co-anchor, fitting the humorous request.

• − The text in the background is slightly clipped.
• − The hand holding the microphone is a bit small and awkwardly placed.

FLUX.1 Kontext [max]

• + Excellent fur texture and individual hair detail
• + Magical, dreamlike lighting with glowing butterflies that accentuates the 'wholesome' vibe
• + Perfectly captures all four animals as requested in the prompt

• − The composition is a bit static with the animals sitting rather than 'tumbling' and 'chasing'
• − The puppy's tongue and mouth area look slightly AI-rendered/smooth

GPT Image 1 Mini

• + Successfully captures the action of 'tumbling' and 'chasing' with dynamic poses
• + Stronger adherence to the 'god rays' lighting effect
• + Included all four requested animals with distinct characteristics

• − The bunny's anatomy is a bit awkward around the neck/ears
• − The animals appear somewhat 'cut out' from the background rather than being fully integrated in the grass

FLUX.1 Kontext [max]

• + Perfectly captures the Studio Ghibli cel-shaded animation style.
• + Preserves the exact composition and poses of the original meme.
• + Excellent use of soft pastel colors and hand-painted watercolor textures in the background.

• − The characters' expressions are slightly softened, losing some of the sharp intensity of the original 'distracted' and 'outraged' looks.

GPT Image 1 Mini

• + Captures a warm, nostalgic mood with a soft-focus atmosphere.
• + Good attention to the textures of the hair and clothing.

• − The style feels more like a colored pencil sketch than Ghibli-inspired animation.
• − The faces are somewhat distorted and lose the specific character of the original photo.
• − Relatively low contrast makes the image feel slightly washed out.

FLUX.1 Kontext [max]

• + Excellent source preservation, maintaining the woman's face and the dog's features accurately.
• + Subtle and realistic blowing hair effect.
• + Successfully added falling leaves as requested.

• − The hand on the right side of the image has mutated into a long, unnatural appendage reaching for the dog.
• − The addition of leaves is a bit sparse compared to the 'energetic' request.

GPT Image 1 Mini

• + Very dynamic wind effect on the hair that looks natural and windswept.
• + Large amount of falling leaves creates a strong sense of lively motion.
• + Generally preserves the overall scene layout well.

• − Significantly altered the woman's facial features, losing the likeness of the source image.
• − The background elements like the bridge and trees have been altered or simplified.
• − Small artifacts around the edges of the leaves and hair.

FLUX.1 Kontext [max]

• + Excellent typography with correct accent placement and spacing.
• + Perfect adherence to the 'light background' and 'warm brown and cream tones' instructions.
• + Authentic vintage texture that looks like a printed emblem.

• − The steam element is very small and slightly disconnected from the cloche knob.

GPT Image 1 Mini

• + Strong iconography for the cloche and steam.
• + Clear layout with high contrast.

• − Failed the negative constraint for a 'light background' by using a black background.
• − The typography is less refined, with an awkward accent mark over the E in 'CAFFÈ'.
• − The color palette leans towards gold and black rather than the requested brown and cream.

FLUX.1 Kontext [max]

• + Excellent typography and high-end graphic design aesthetic.
• + Includes highly accurate silhouettes of the crew with correct names.
• + Effective use of the requested NASA-inspired color palette and subtle gradients.

• − Failed to include exactly 6 steps as requested in the prompt.
• − The rocket icon and some lunar icons are more generic/abstract than Saturn V or Lunar Module specifically.
• − Text logic is confusing with labels pointing to the wrong visual elements.

GPT Image 1 Mini

• + Followed the specific 6-step sequence exactly as outlined in the prompt.
• + Icons for the Saturn V and Lunar Module are much more accurate and consistent.
• + Clean, consistent flat-vector illustration style with a clear logical flow.

• − The 'Translunar' trajectory icon is a bit messy and over-complicated compared to other icons.
• − Text rendering is slightly less refined than Model A.
• − Composition feels slightly cramped at the top.