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

FLUX.1 Kontext [dev]

• + Excellent photographic realism and lighting
• + Strong adherence to the spatial arrangement
• + High resolution with realistic textures on the book and plant

• − The glass cube has a solid base that looks more like a mirror than transparent glass

GPT Image 1 Mini

• + Successfully placed all elements according to the prompt
• + Clean, minimalist composition
• + Accurate representation of a fully transparent glass cube

• − The plant is blurred and not clearly 'behind the cube' in a way that shows through the glass
• − Lighting is a bit flat compared to Model A

FLUX.1 Kontext [dev]

• + Strong adherence to the 'reflections on wet pavement' and 'rain' prompts
• + Excellent clarity and color vibrantness
• + Successfully captures the busy Japanese street atmosphere

• − The subject is posing with the bike rather than repairing it
• − The composition feels very centered and clean, missing the 'imperfect framing' request
• − The man's skin looks a bit too smooth for 'natural skin texture'

GPT Image 1 Mini

• + Perfect adherence to the 'repairing' action and 'natural skin texture'
• + Captures the 'imperfect framing' and 'candid' feel much better
• + Realistic skin tones and weathered details on both the man and the bike

• − The 'light rain' and 'wet pavement' reflections are significantly more subtle than in Model A
• − The car in the background lacks noticeable 'motion blur' as requested

FLUX.1 Kontext [dev]

• + Strong cinematic lighting with high contrast
• + Highly detailed engraving on the breastplate
• + Clear bokeh sparks in the background

• − Missed the request for braided hair with beads
• − Skin texture looks slightly overly smoothed or airbrushed compared to Model B
• − Faint scars look more like fresh red marks

GPT Image 1 Mini

• + Excellent adherence to the 'braided hair' and 'battle-worn' descriptor with realistic dirt and scars
• + Superior texture on leather straps and metal surfaces
• + More lifelike, weathered facial features

• − The lighting is a bit more diffused than 'warm torchlight' suggests
• − The bokeh sparks are a bit more uniform and less dynamic than in Image A

FLUX.1 Kontext [dev]

• + Excellent high-fidelity food photography with realistic textures.
• + Dynamic and creative layout with alternating text and image blocks.
• + Sophisticated use of negative space and typography scale.

• − Text is largely gibberish and contains visual artifacts inside letters.
• − The grid is somewhat irregular rather than a clear, functional menu list.

GPT Image 1 Mini

• + Perfect text rendering for section headers like 'PIZZA' and 'MAINS'.
• + Highly functional and logical menu layout that is easy to read.
• + Clean grid alignment that directly follows the prompt's structural requirements.

• − Food photography looks slightly more artificial and clipart-like compared to Model A.
• − Composition is a bit generic and lacks the 'modern designer' flair of Model A.

FLUX.1 Kontext [dev]

• + Excellent photographic quality and vivid colors.
• + Strong, clear typography for the main title.
• + Clean and polished graphic design aesthetic.

• − Failed to provide an 'exploded' view; the burger is fully assembled.
• − Contains a spelling error in the secondary text: 'LNHLY'.
• − The starburst is a flat graphic rather than a fiery effect.

GPT Image 1 Mini

• + Followed the 'exploded burger' instruction perfectly.
• + Accurately rendered all requested fiery text effects and glowing starburst.
• + Perfect spelling and adherence to all prompt elements.

• − The lighting on the burger feels slightly flatter compared to Model A.
• − Overall image has a bit more digital noise/grain intended for a 'fiery' feel.

FLUX.1 Kontext [dev]

• + Natural variation in handwriting styles across the board
• + Strong chalk texture realism
• + Excellent framing and wood grain on the chalkboard

• − Numerous spelling errors including 'Mashroom', 'Risoktso', and 'Octpus'
• − Muddled and illegible Month/Date text
• − Repetitive words in sentences

GPT Image 1 Mini

• + Perfect spelling of all menu items and dates
• + Very clean and readable layout
• + Excellent chalk texture within the letters

• − Text looks somewhat like a digital font despite being requested as handwritten
• − Lower contrast between the chalk and the board compared to model A
• − Slightly less 'natural' handwriting flow

FLUX.1 Kontext [dev]

• + Successfully followed the difficult positional instruction of the horse riding the astronaut.
• + High detail on the astronaut's face and suit textures.
• + Clean, modern cinematic aesthetic with sharp focus.

• − The horse's hind legs are poorly integrated into the astronaut's shoulders.
• − The scale of the horse relative to the man feels slightly inconsistent.

GPT Image 1 Mini

• + Excellent atmospheric lighting and background starfield.
• + Great sense of motion in the horse's mane and pose.

• − Failed the core prompt instruction to have the horse on top.
• − Standard cliché interpretation instead of the requested surreal reversal.

FLUX.1 Kontext [dev]

• + Excellent photorealism in the human passenger's face and hands
• + High clarity and sharp details on the capybara's fur and clothing

• − One paw is on his lap instead of both on the steering wheel as requested
• − The capybara's face looks slightly hybridized with a bear or dog

GPT Image 1 Mini

• + Perfect capybara anatomy and distinctive facial structure
• + Better lighting and bokeh effect in the background Manhattan skyline
• + Captures the classic 'taxi driver' cap style with the checkered band

• − The passenger's face is slightly out of focus and less realistic than the driver
• − Only one paw is clearly visible on the steering wheel

FLUX.1 Kontext [dev]

• + Strong, glowing central jack-o-lantern
• + Clean, readable main title text

• − Garbled and unreadable text in the scroll banner
• − Incorrect spelling for the event location ('ARGIIVAH'S')
• − The background and elements feel like separate clip-art rather than a cohesive cinematic scene

GPT Image 1 Mini

• + Perfect text rendering for all requested fields, including the banner and address
• + Superior cinematic atmosphere with a textured parchment feel
• + Highly detailed border containing webs and thorns as requested

• − The lighting on the pumpkin is slightly more muted compared to the glow in Model A

FLUX.1 Kontext [dev]

• + Excellent preservation of the original jacket and background.
• + Matches the beard color and texture to the new head hair reasonably well.

• − Significantly alters the facial structure, making the person look like a different individual.
• − The hair looks somewhat stiff and 'helmet-like' with a very straight, artificial hairline.
• − Unnecessarily changed the frame color and thickness of the glasses.

GPT Image 1 Mini

• + Maintains the subject's identity and facial features much better than the competitor.
• + The hair has a more natural, messy texture that fits the outdoorsy setting.
• + Preserves the original eyewear and clothing accurately.

• − There is some slight blurring/softening of detail around the hairline.
• − The hair density at the crown looks a bit clumped.

FLUX.1 Kontext [dev]

• + Very clean, bold typography.
• + Solid, even light blue background as requested.
• + Simple and distinct 3D miniature style.

• − The flag icon is incorrect and abstract.
• − The sushi design is oversimplified and looks more like a pill or toy than a recognizable food item.
• − Isometric angle is a bit flat compared to the prompt's 45-degree request.

GPT Image 1 Mini

• + Excellent adherence to the 45-degree isometric perspective.
• + Includes an accurate Japanese flag icon.
• + Beautiful textures on the sushi and wooden diorama base that fit the 'refined PBR' request.

• − The text layout is slightly less 'bold' and impactful than Model A.
• − Includes chopsticks which were not explicitly requested, though they fit the scene well.

FLUX.1 Kontext [dev]

• + Successfully incorporates the TV show anchor element with a monitor and desk.
• + Retains the denim shirt and general facial structure of the subject.
• + Captures the request for a dog.

• − Entirely misses the hockey requirement.
• − The 'JOB' text is redundant and non-creative.
• − The 'caricature' style is more of a generic cartoon/comic illustration than an exaggerated caricature.

GPT Image 1 Mini

• + Successfully incorporates all requested elements: anchor profession, dog, and hockey equipment.
• + The caricature style is much more effective with properly exaggerated facial features.
• + Stronger composition that tells a cohesive story of the subject's life.

• − The hand holding the microphone has some slight anatomical oddities.
• − Text on the hockey puck is absent/blurry.

FLUX.1 Kontext [dev]

• + Excellent dynamic lighting with a warm, glowing atmosphere
• + High detail in the grass and surrounding wildflowers
• + Captured the 'playful chasing' action well with convincing poses

• − Failed to include all requested animals, missing the rabbit and fox kit
• − The cats look more like two versions of a kitten rather than a tabby kitten and the other requested animals
• − The middle kitten's paws appear somewhat blurred and anatomically awkward

GPT Image 1 Mini

• + Perfect adherence to the prompt, including all four specific animals: puppy, tabby kitten, bunny, and fox kit
• + Beautifully rendered fur textures and expressive eyes that match the 'hyper-photorealistic' request
• + Superior implementation of god rays and dew sparkles in the grass

• − The fox kit has slightly dark, muddy-looking front paws
• − Minor anatomical oddity on the kitten's tail which appears very long and thick

FLUX.1 Kontext [dev]

• + Excellent preservation of the original composition and poses.
• + Captures the clean line-art style and color palette of modern anime.
• + Successfully brings the out-of-focus foreground woman into clear focus while maintaining her appearance.

• − The aesthetic leans more toward generic modern anime than the specific hand-painted Studio Ghibli look.
• − Characters lose some of the unique facial expressions from the source.

GPT Image 1 Mini

• + Perfectly captures the Studio Ghibli 'hand-painted' texture and soft pastel aesthetic.
• + Maintains the emotional intensity of the faces, especially the man's expression.
• + Excellent watercolor-like lighting and warm, nostalgic atmosphere.

• − The overall image is slightly more yellowed and blurred compared to the source.
• − Slight loss of structural detail in the background buildings.

FLUX.1 Kontext [dev]

• + Excellent source preservation, maintaining the woman's face and the dog's features almost perfectly.
• + Subtle and realistic hair motion that looks natural.
• + Minimal artifacts and clean background preservation.

• − The 'flying leaves' are barely noticeable and very small.
• − Does not feel as 'energetic' as the prompt requested.

GPT Image 1 Mini

• + Successfully adds a large volume of flying leaves, creating a very dynamic atmosphere.
• + Good interpretation of 'energetic and lively' with significant hair movement.

• − Poorer preservation of the woman's facial features and the dog's face compared to the original.
• − Introduces noticeable artifacts around the dog's leash where it meets the hand.
• − The leaves appear somewhat flat and over-saturated.

FLUX.1 Kontext [dev]

• + Excellent typography with perfect spelling and accent marks.
• + Clean vector-style execution suitable for a professional logo.
• + Perfectly adheres to the light background and warm brown tone requirement.

• − The cloche dome icon is a bit abstract and looks slightly more like a building or crown.
• − Misses the specific 'banner' element for the date.

GPT Image 1 Mini

• + Included the requested banner for the 'Est. 1720' text.
• + The cloche dome illustration is more recognizable and features a nice 'etched' texture.
• + Captures the vintage feel well with serif typography.

• − Failed the 'light background' prompt requirement by using a black background.
• − Typography has minor aliasing and lacks the clean vector crispness of Model A.

FLUX.1 Kontext [dev]

• + Strong NASA-inspired color palette with navy and muted red.
• + Good use of flat vector aesthetics with subtle gradients.
• + Creative interpretation of iconic landing silhouettes.

• − Significant spelling errors throughout the text (e.g., 'Apolo', 'Descnt').
• − The iconography is messy and abstract, failing to clearly represent the requested steps like the Saturn V.
• − Numerical ordering is inconsistent and repetitive.

GPT Image 1 Mini

• + Excellent adherence to the sequential steps requested in the prompt.
• + Crisp, clean vector lines and highly legible typography.
• + Accurate and recognizable iconography for the Saturn V and Lunar Module.

• − Translunar trajectory arc is a bit looped and illogical visually.
• − The layout feels slightly sparse compared to a full poster.