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

FLUX.1 [dev]

• + Sophisticated glass rendering with realistic internal reflections and refraction.
• + Exhibits high photographic quality with naturalistic wood grain and soft bokeh.
• + Accurately captures the lighting source from the left as requested.

• − The plant is positioned more to the side than 'behind' compared to the other model.
• − The sphere appears to be floating mid-air rather than resting on the bottom of the cube.

GPT Image 1

• + Follows the spatial arrangement perfectly, with the plant clearly visible behind the glass.
• + Clear, logical geometry in the composition of the cube and book.
• + Correctly interprets the sphere's placement inside the cube.

• − The glass cube is rendered with thick, teal-tinted edges that look a bit artificial.
• − Less impressive handling of complex refractions through the glass compared to FLUX.1 [dev].

FLUX.1 [dev]

• + Excellent handling of wet pavement reflections and light rain particles
• + Cinematic composition with a realistic 50mm lens feel
• + Highly detailed clothing textures and realistic skin

• − The man is holding the handlebars rather than performing a repair
• − The bicycle design is a bit modern and generic for the 'candid' prompt

GPT Image 1

• + Strong adherence to the 'repairing' action with a crouching pose
• + Captures the 'imperfect framing' and 'candid' feel more effectively
• + Very natural, non-stylized skin texture on the man's face

• − The bicycle frame geometry is physically nonsensical near the rear hub
• − Less noticeable motion blur on passing cars compared to Image A

FLUX.1 [dev]

• + Excellent photographic skin texture and lifelike eyes
• + Vibrant and high-quality bokeh effect
• + Clean and symmetrical composition

• − Missed the 'small beads' in the hair braids
• − The character looks too pristine and model-like for the 'battle-worn' prompt
• − Plate armor is basic rather than 'ornate engraved'

GPT Image 1

• + Includes small beads in the braids and ornate engravings on the armor
• + Character has a realistic 'battle-worn' expression with dirt and grime
• + More dramatic and atmospheric warm lighting

• − Slightly lower skin texture resolution compared to Model A
• − Composition is a bit tighter, cutting off some detail
• − Armor texture looks slightly more digital than metallic

FLUX.1 [dev]

• + Excellent professional graphic design layout with whitespace
• + Includes a clean restaurant branding area
• + High quality, integrated food photography

• − Text consists largely of jibberish/placeholder characters
• − Relies on a diagonal layout rather than a standard grid

GPT Image 1

• + Successfully follows the requested grid layout for food photos
• + Text is legible and uses bold sans-serif fonts as requested
• + Includes specific sections for Appetizers and Pizza

• − Lacks the 'Mains' category section requested in the prompt
• − Slightly less 'professional' or 'modern' aesthetic compared to Model A
• − Text description is repetitive and contains minor typos like 'descrigion'

FLUX.1 [dev]

• + Excellent photorealistic texture on the meat patties and vegetables
• + Clean, professional composition with a focused lighting effect

• − Completely failed to include the primary 'MAGIC BURGER' text
• − Failed to render the price in a starburst as requested
• − The explosion effect is very vertical and static, lacking dynamic motion

GPT Image 1

• + Successfully integrated all requested text including the fiery effect
• + Perfectly followed the instruction for the starburst price tag
• + Dynamic composition with embers and liquid splashes creating a sense of motion

• − The price '€.99' is slightly incorrect, missing the '6'
• − The 'MAGIC BURGER' text is cut off at the top of the frame

FLUX.1 [dev]

• + Excellent text rendering with no spelling errors.
• + Beautiful high-contrast composition and realistic café background.
• + Followed the prompt to complete the third menu item creatively with cookies.

• − The 'chalk' texture looks a bit more like a digital marker or smooth paint than dry chalk.
• − The handwriting style for the body text looks slightly too uniform, like an 'ink' font.

GPT Image 1

• + Exceptional realistic chalk texture, including the grainy edges of the strokes.
• + The title font feels more authentic to a handwritten chalkboard style.

• − Missing the dollar sign for the last item (Cookies 9).
• − Lower visual contrast compared to the first image, making it look a bit dull.
• − The letter 'i' in 'Risotto' looks slightly malformed.

FLUX.1 [dev]

• + Features a smooth, ethereal lighting that fits the cinematic request.
• + The white horse and suit create a cohesive color palette.

• − Completely failed the semantic challenge of 'horse on top' of the astronaut.
• − Anatomical issues with the horse's legs, including a strange double-jointed appearance on the hind leg.

GPT Image 1

• + Excellent texture work on the space suit and horse hair.
• + Higher level of detail in the background with visible planets and nebulae.

• − Failed to follow the specific instruction to put the 'horse on top' of the astronaut.
• − The composition is a cliché interpretation of the 'space cowboy' trope despite the prompt's push for surrealism.

FLUX.1 [dev]

• + High resolution and crisp rendering of the capybara's fur.
• + Very aesthetically pleasing lighting and bokeh effect.
• + Accurately represents the 'bored' expression requested for the passenger.

• − The passenger is sitting in the front passenger seat rather than the back seat as requested.
• − The capybara's paws are not placed correctly on the steering wheel, appearing to float or hold nothing.
• − The steering wheel is on the right side, which is incorrect for a New York taxi.

GPT Image 1

• + Accurately places the passenger in the back seat as requested.
• + Correctly places both paws of the capybara on the steering wheel.
• + The taxi driver cap is much more authentic to a professional uniform.

• − The passenger's face is slightly out of focus and less detailed.
• − The steering wheel is also on the right side, which contradicts the New York setting.
• − The overall image is slightly darker and less vibrant than the competitor.

FLUX.1 [dev]

• + Strong cinematic lighting with a vibrant glowing effect
• + Intricate thorny border framing the entire composition
• + High contrast and sharp, clean illustration style

• − Several text hallucinations including 'Falloween Rantcy' and 'You Tre'
• − Formatting errors in the event details such as '30, 10, 2026' and repeated time tags

GPT Image 1

• + Perfect adherence to the requested text for the header and subtitle banner
• + Captures the 'vintage gothic' aesthetic well with muted tones and parchment texture
• + Includes both webs and thorns in the border as requested

• − The location 'The Arches' is incorrectly mapped to the 'TIME' label
• − Visuals are a bit muddy and lower contrast compared to Model A

FLUX.1 [dev]

• + Excellent soft-focus photography aesthetic
• + Beautifully rendered translucent textures on the fish
• + Accurate 45-degree isometric projection

• − Text rendering is poor with spelling errors like 'SUSH CATON'
• − Background/text relationship is cluttered by the flowers
• − Contrast is a bit low

GPT Image 1

• + Perfect text rendering for both 'JAPAN' and 'SUSHI'
• + Strong '3D cartoon' aesthetic as requested
• + Clean, bold composition with high clarity

• − The diorama base is a bit large relative to the plate
• − Materials look slightly more like clay than 'realistic PBR'

FLUX.1 [dev]

• + Features soft, vibrant colors that enhance the 'wholesome' vibe.
• + Clean composition with distinct, well-placed characters.

• − Fails the prompt's 'hyper-photorealistic' requirement, looking more like a 3D animation or digital illustration.
• − Missing the kitten entirely, and depicts one animal as a hybrid squirrel/rabbit creature.
• − Characters are posed in a static line rather than 'tumbling together'.

GPT Image 1

• + Excellent adherence to 'hyper-photorealistic' with realistic fur textures and anatomy.
• + Accurately includes all four requested animals: golden retriever, tabby kitten, bunny, and fox kit.
• + Dynamic composition captures the 'chasing' and 'tumbling' action requested.

• − The fox kit has three front paws visible, indicating a minor anatomical artifact.
• − The butterfly in the top left corner is a bit large in scale compared to the animals.

FLUX.1 [dev]

• + Elegant layout with a professional vector emblem feel
• + Good use of negative space and classic typography flourishes

• − Significant spelling errors including 'Flarilaan' and 'Reseaurant'
• − Includes hallucinated numbers ('11011', '1941') not present in the prompt

GPT Image 1

• + Flawless text rendering exactly matching the prompt including 'Caffè'
• + Strong adherence to the texture requirement and minimalist vector style
• + Perfectly placed banner and cloche icon

• − Interpreted 'light background' as a dark background (though the asset itself is brown/cream)
• − Minimalist design might feel slightly simple compared to traditional emblems

FLUX.1 [dev]

• + Elegant layout with a professional infographic feel.
• + Strong adherence to the flat-vector style with sophisticated color blending.
• + Excellent visual balance and clean iconography.

• − Text consists of gibberish despite being legible.
• − Fails to follow the logical sequence of steps requested in the prompt.

GPT Image 1

• + Excellent text rendering with correct names and labels.
• + Successfully includes almost all requested technical steps and icons.
• + Clean, communicative design that functions as an actual infographic.

• − Simple composition that feels slightly cluttered compared to model A.
• − Minor spelling error with 'EARLLUNAR' in the bottom label.