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

FLUX.1 Kontext [pro]

• + Excellent adherence to lighting and placement instructions.
• + Highly realistic glass textures with clean edges.
• + Great spatial awareness with the blue sphere resting naturally on the bottom surface.

• − The plant is significantly out of focus compared to the rest of the scene.
• − The blue sphere has a slightly fuzzy or felt-like texture rather than appearing perfectly smooth.

GPT Image 1

• + Strong color saturation and sharp details on the book and plant.
• + Accurate material representation for the glass cube with slight green-tinted edges.
• + Clear visibility of the plant through the glass as requested.

• − The lighting feels slightly more omnidirectional than the requested 'soft light from the left'.
• − The bottom of the cube has an unexpected metallic reflective base rather than being pure glass.

FLUX.1 Kontext [pro]

• + Natural skin texture and realistic face rendering.
• + Good application of rain effects and wet pavement reflections.
• + Accurate shallow depth of field for a 50mm look.

• − The subject appears to be riding or holding the bike rather than repairing it.
• − The bicycle handlebars and cables show some structural AI artifacts.

GPT Image 1

• + Strong adherence to the 'repairing' action in the prompt.
• + Excellent composition with a side profile that feels candid.
• + Realistic textures on the bicycle seat and wet metal.

• − The hands merging into the bike chain area are a bit messy.
• − Less noticeable motion blur from passing cars compared to the other model.

FLUX.1 Kontext [pro]

• + Excellent photorealistic skin texture and lifelike eyes
• + Beautifully detailed engraving on the plate armor
• + Subtle and realistic integration of the beard and facial hair

• − The beads in the hair are very minimal and less noticeable than requested
• − Lighting is a bit flat compared to the dramatic 'torchlight' prompt

GPT Image 1

• + Superior adherence to the 'hair braided with small beads' prompt
• + Very gritly, battle-worn appearance with convincing dirt and scars
• + Dramatic warm lighting that effectively highlights the metal textures

• − Skin texture is slightly more stylized/painterly than Model A
• − The pupils in the eyes look slightly inconsistent in shape

FLUX.1 Kontext [pro]

• + Excellent typography with clean, bold sans-serif fonts.
• + Sophisticated layout that feels like a real graphic design piece.
• + Accurate interpretation of the sections requested including categorized food items.

• − The text content is mostly gibberish.
• − The food photos are slightly less vibrant compared to Model B.

GPT Image 1

• + Features a very clean 2x2 grid of high-quality food photography.
• + Text labels are closer to English words with logical price formatting.
• + High-contrast colors and professional lighting in the images.

• − The 'Mains' section is poorly integrated into the layout.
• − Composition is a bit generic and lacks the 'graphic design' flair of Model A.
• − Some spelling errors like 'descrigion' and 'Apperoiation'.

FLUX.1 Kontext [pro]

• + Excellent photographic quality and lighting on the burger components.
• + Strong typography rendering with clear, glowing effects.
• + Creates a highly professional advertisement layout with a rich background.

• − Repeated the price three times in the image unnecessarily.
• − Missing the starburst element for the price as requested.
• − The 'exploded' effect is less pronounced than requested, as the burger is mostly stacked.

GPT Image 1

• + Perfect adherence to the 'exploded' instruction with clear gaps between all layers.
• + Includes the starburst element for the price as specifically requested.
• + Consistent fiery text effect across all copy.

• − Failed the price accuracy, showing '€.99' instead of '€6.99'.
• − The bottom bun looks slightly flat and untextured compared to the rest of the image.
• − The overall composition feels a bit more cramped at the top.

FLUX.1 Kontext [pro]

• + Excellent chalk texture with realistic dusty strokes
• + High prompt adherence for specific menu items and date
• + Logical chalkboard layout with a wooden frame

• − Significant spelling errors in the footer text at the bottom
• − The title is not in 'elegant cursive' as requested

GPT Image 1

• + Perfect spelling in the footer text
• + Consistent handwriting style throughout the entire board
• + Clean and centered composition

• − Failed to include the '$' sign for the final price item
• − Calculated chalk texture looks slightly like a digital brush rather than physical chalk
• − The title is printed caps rather than elegant cursive

FLUX.1 Kontext [pro]

• + Excellent adherence to the complex spatial requirement of the horse being on top.
• + Creative depiction of a smaller astronaut mounting the horse which is mounting the larger astronaut.
• + Clean, cinematic lighting and high-quality textures on the spacesuit.

• − The anatomy of the horse's rear legs and the small astronaut's merging are a bit confusing.

GPT Image 1

• + High visual quality with a classic cinematic aesthetic.
• + Strong composition and clear subject rendering.

• − Failed the core prompt instruction of having the horse on top of the astronaut.
• − Standard interpretation lacking the requested surreal role-reversal.

FLUX.1 Kontext [pro]

• + Excellent fur texture rendering
• + Deep vibrant colors in the taxi paint
• + Highly realistic cinematic lighting

• − The passenger is on a phone call instead of looking at the phone as requested
• − Only one paw is clearly visible on the steering wheel
• − The perspective is awkward, looking in through the driver side window

GPT Image 1

• + Perfect adherence to the passenger's action and expression
• + Stronger composition with a front-on view of the driver and passenger
• + Accurately shows both paws on the steering wheel as requested

• − The capybara's hands look slightly more humanoid/primate-like than paws
• − Overall image is a bit darker and less sharp than Model A

FLUX.1 Kontext [pro]

• + Excellent typography style that matches the 'gothic' request perfectly.
• + Very high resolution details in the pumpkin and moon.
• + Strict adherence to the banner text and date format.

• − Includes a line of gibberish text in the event details section.
• − The layout is a bit bottom-heavy with small text.

GPT Image 1

• + Better 'parchment' texture and color palette for a vintage look.
• + Organized and clean text layout that is easier to read.
• + Accurate rendering of the requested event details without hallucinated lines.

• − Text is in a serif font rather than a truly 'elegant gothic' style.
• − Jack-o-lantern and background elements are slightly blurry or painterly compared to Model A.

FLUX.1 Kontext [pro]

• + Excellent preservation of the original facial features and bone structure
• + Realistic hair texture and natural integration with the sideburns
• + Maintains identical lighting and background to the source image

• − The forehead appears slightly shortened compared to the original

GPT Image 1

• + Successfully adds a thick head of hair
• + Preserves the overall composition of the image

• − Significantly alters the person's face, making them look like a different individual
• − The hair texture looks somewhat like a wig and lacks a natural-looking hairline
• − The glasses and eye area were altered unnecessarily

FLUX.1 Kontext [pro]

• + Excellent typography style that matches the 3D aesthetic.
• + High-quality texture on the salmon and rice, showing great PBR material properties.
• + Clean and minimalist composition that looks like a professional 3D render.

• − The flag icon is integrated oddly into the text line rather than being a separate element.
• − The rice grains look slightly larger than natural, resembling pearls.

GPT Image 1

• + Perfect adherence to the layout instructions, including the flag icon placement.
• + More variety in the sushi types which makes for a more interesting scene.
• + Clean, soft lighting that perfectly matches the 'cartoon miniature' request.

• − Slightly less realistic PBR textures on the salmon compared to the other model.
• − The flag icon is a flat 2D graphic which slightly clashes with the 3D text styling.

FLUX.1 Kontext [pro]

• + Successfully converts the subject into a stylized vector-art caricature.
• + Strongly preserves the clothing and selfie-pose composition from the source image.

• − Fails completely to incorporate the requested job (anchor), dogs, or hockey themes.
• − The edit is too simple, focusing only on the style change rather than the content additions.

GPT Image 1

• + Excellent adherence to all prompt details, including the news desk, dog, and hockey elements.
• + Captures the classic 'big-head' caricature style with high-quality watercolor texture.
• + Maintains the subject's likeness and clothing while transforming the setting.

• − The hands holding the papers are slightly malformed.
• − Text on the hockey stick is incoherent.

FLUX.1 Kontext [pro]

• + Excellent depiction of warm golden light and god rays.
• + Includes all four requested animals with very soft, fine fur textures.

• − The characters are mostly sitting rather than 'playfully chasing and tumbling'.
• − The rabbit lacks defining bunny features, looking more like a small kitten-rabbit hybrid.

GPT Image 1

• + Strong prompt adherence to the action, showing the animals actually running and chasing.
• + Individual animal species are more distinctly recognizable, especially the rabbit and fox.
• + Better composition that feels dynamic and joyful.

• − The fox kit has three front paws visible due to a developmental artifact.
• − The lighting lacks the specific 'god rays' mentioned in the prompt compared to the other model.

FLUX.1 Kontext [pro]

• + Excellent preservation of the original image's composition and character poses.
• + Captures the iconic Ghibli character design style perfectly while staying recognizable.
• + Hand-painted watercolor texture is very high quality and consistent.

• − The man's expression is slightly more neutral/melancholic than the original pucker/surprise.

GPT Image 1

• + Strong implementation of the requested 'deeply dreamy' and 'vibrant pastel' color palette.
• + Good use of soft lighting and hazy textures.

• − Loss of significant source detail, specifically the sharp plaid pattern on the shirt.
• − The woman in the foreground has her eyes closed, losing the gaze towards the camera.
• − Character faces are a bit more generic and less polished than model A.

FLUX.1 Kontext [pro]

• + Excellent preservation of the woman's face and original features.
• + Highly realistic hair-blowing effect that looks naturally integrated.
• + Subtle and tasteful addition of flying leaves that do not clutter the frame.

• − The dog's tail position changed significantly from the original source.

GPT Image 1

• + Stronger 'energetic' feel with more numerous leaves scattered throughout the scene.
• + Effective hair motion that suggests a gust of wind.

• − Noticeable anatomical distortion on the woman's left hand (holding the leash).
• − The leash itself has become disconnected and fragmented with a strange loop.
• − The image has a more 'photoshopped' or artificial look compared to the source.

FLUX.1 Kontext [pro]

• + Excellent typography with a correct grave accent on 'Caffè'.
• + Beautifully rendered vintage paper texture that matches the 'subtle texture' prompt.
• + Highly balanced composition with a classic vector emblem feel.

• − The steam element is a bit thick, resembling a leaf or solid shape more than vapor.

GPT Image 1

• + Strong contrast with the use of warm brown and cream tones.
• + Clean iconography for the cloche and steam elements.
• + Accurate text rendering for the restaurant name and date.

• − Ignored the 'light background' instruction, using a solid black background instead.
• − Incorrect accent mark on 'Caffè' (it uses a grave accent in the prompt and real life, but the model used an acute-style mark).

FLUX.1 Kontext [pro]

• + Excellent high-resolution aesthetics with a polished 'modern vector' feel.
• + Creative use of a trajectory path to connect different stages of the mission.
• + Successfully captures a NASA-inspired color palette and atmospheric depth.

• − Confused logical mapping, such as labeling a ringed planet (Saturn) as 'Earth'.
• − Fails to provide the specific 1-6 numbered steps requested in the prompt.
• − Poor text legibility and nonsensical labels like '2, f ar Collins'.

GPT Image 1

• + Strict adherence to the iconography requested, including the specific lunar module icons.
• + Better organized layout that resembles a standard infographic with distinct sections.
• + Higher degree of text accuracy for the labels and crew names.

• − Significant spelling errors such as 'EARLLUNAR' and inconsistent font spacing.
• − Visual style is slightly more dated/clunky than the requested modern aesthetic.
• − Missing the 'Lunar Orbit' step specifically requested, jumping from Translunar to Descent.