FLUX1.1 [pro] Black Forest Labs vs GPT Image 1 Mini OpenAI

FLUX1.1 [pro]

• + Exquisite lighting and reflections on the glass and sphere surfaces.
• + Realistic botanical background with accurate refraction through the glass.
• + High-resolution texture on the red book cover.

• − The glass container is a tall rectangular prism rather than a 'cube'.
• − The sphere appears to be floating unnaturally in the center.

GPT Image 1 Mini

• + The glass container is a perfect cube as requested.
• + Realistic placement of the sphere resting on the bottom surface.
• + Accurate interpretation of 'small blue sphere' relative to the cube size.

• − The green plant is placed to the side rather than 'behind the cube' visible through the glass.
• − The lighting on the book is somewhat flat compared to the rest of the scene.

FLUX1.1 [pro]

• + Excellent wet pavement reflections that enhance the cinematic feel.
• + Good use of bokeh and background lighting to create depth.
• + Strong adherence to the 'light rain' and 'passing cars' prompts.

• − Major anatomical and structural issues with the bicycle's frame and the man's pose.
• − The bicycle appears fragmented and physically impossible under the handlebars.
• − The man's hands on the handlebars do not look natural.

GPT Image 1 Mini

• + Superb skin texture and facial detail that looks highly realistic.
• + The bicycle's structure remains coherent and logical while being repaired.
• + Captures the 'candid' and 'natural' look perfectly with subtle rain droplets on surfaces.

• − Misses the 'motion blur from passing cars' request in the background.
• − The background cars are blurred by depth of field but don't clearly show motion.

FLUX1.1 [pro]

• + Extremely lifelike eye details with realistic light refraction.
• + Superior skin texture showing pores, dirt, and fine scarring with high fidelity.
• + Excellent depth of field with realistic foreground and background blur.

• − Fails to show the 'small beads' requested in the braids.
• − The engraving on the armor is less intricate than the competing image.

GPT Image 1 Mini

• + Beautifully intricate and coherent engraving on the plate armor.
• + Captures the warm torchlight ambiance very effectively across the whole scene.
• + Shows visible beads within the braided hair as requested.

• − Skin texture looks somewhat painterly or smoothed compared to Model A.
• − The eyes look slightly flat and lack the 'lifelike' sparkle of a high-end photograph.

FLUX1.1 [pro]

• + Excellent photo-realistic food photography with authentic lighting
• + Captures the professional look of a tri-fold or complex menu layout
• + Includes actual price markers and descriptive text elements

• − The text is mostly gibberish upon close inspection
• − Grid for food photos is somewhat messy and inconsistent in cell sizing

GPT Image 1 Mini

• + Perfectly adheres to the grid request with clean, uniform photo blocks
• + Clear, legible section headers for Appetizers, Pizza, and Mains
• + Very clean, minimalist aesthetic that fits the modern prompt

• − Lacks actual menu item descriptions or prices under the headers
• − The food images appear slightly more generic or stock-like compared to Model A

FLUX1.1 [pro]

• + Excellent photorealistic texture on the burger patties and buns
• + Good cinematic lighting and high-quality background bokeh

• − Failed to explode the burger layers, resulting in a mostly intact stack
• − Text lacks the requested 'fiery, glowing effect' and is doubled/poorly arranged
• − Missing the starburst for the price

GPT Image 1 Mini

• + Strong prompt adherence for the 'exploded' architecture of the burger
• + Perfectly followed text instructions, including the fiery effect and starburst price tag
• + Colors are vibrant and evoke a magical/fiery theme well

• − The burger ingredients look slightly less realistic compared to Model A
• − Composition is a bit static despite the floating elements

FLUX1.1 [pro]

• + Excellent handwritten flow with natural slants and variations.
• + Beautifully rendered background with cozy café atmosphere and bokeh lighting.

• − Hallucinated text by splitting the octopus item into two separate lines with incorrect prices.
• − Failed to complete the final word correctly, rendering 'Chipkies' instead of 'Cookies'.

GPT Image 1 Mini

• + Perfect text accuracy, including the spelling of 'Cookies' and all price values.
• + Excellent chalk texture that looks authentically dusty and porous.
• + Successfully followed the layout for long menu items by wrapping text properly.

• − The cursive title is very blocky and lacks the 'elegant cursive' style requested.
• − The background is very plain compared to the requested 'cozy café' atmosphere.

FLUX1.1 [pro]

• + Excellent cinematic lighting and color range.
• + High level of detail in the astronaut's suit and horse's mane.

• − Failed the negative constraint; the astronaut is riding the horse, not the other way around.
• − The horse has an extra leg visible at the bottom.

GPT Image 1 Mini

• + Atmospheric space background with good depth.
• + Clean composition and texture on the horse.

• − Failed the negative constraint; the astronaut is riding the horse.
• − Slightly less creative interpretation of the cinematic prompt compared to the lighting in model A.

FLUX1.1 [pro]

• + Dynamic lighting with warm yellow tones consistent with taxi interior
• + High aesthetic quality and sharp focus on subjects

• − The passenger is sitting in the front passenger seat rather than the back seat
• − The capybara's paws are not visible on the steering wheel
• − The perspective is from the trunk or rear window rather than inside the cabin viewing the driver from the back

GPT Image 1 Mini

• + Strict adherence to the layout with the passenger correctly placed in the back seat
• + Accurately depicts the capybara with its paws on the steering wheel
• + Perfectly captures the 'bored' expression of the businesswoman

• − Lighting is slightly flatter and less cinematic than the competitor
• − The passenger's right hand holding the phone has some minor anatomical blurring

FLUX1.1 [pro]

• + Features a vivid, high-contrast glow on the jack-o-lantern and moon
• + Intricate thorn border and jagged paper edges enhance the gothic theme
• + Excellent cinematic lighting and depth of field

• − Significant text repetition and spelling errors ('Halloween Party Party', 'You inivted')
• − Does not adhere to the requested square format
• − Layout feels cramped with redundant lines of information at the bottom

GPT Image 1 Mini

• + Perfect text accuracy and clean typography layout
• + Adheres correctly to the square format requested
• + Authentic vintage parchment texture and moody gothic aesthetic

• − Visuals are slightly more muted and less 'cinematic' than Model A
• − The border details (webs and thorns) are quite dark and difficult to see

FLUX1.1 [pro]

• + Excellent variety of sushi items on the diorama
• + High-quality soft clay-like textures with clean shading
• + Perfectly matches the 45-degree isometric angle requested

• − Failed to include the small flag icon
• − The 'SUSHI' text is somewhat thin and harder to read compared to 'JAPAN'

GPT Image 1 Mini

• + Successfully included all text elements and the Japanese flag icon
• + Distinct and clear font choice for both words
• + Very clean, minimal composition that follows 'minimal garnish' instruction

• − The diorama base has a slight perspective misalignment with the round plate
• − Lower complexity in the sushi variety compared to Model A

FLUX1.1 [pro]

• + Excellent rim lighting and backlighting effect on the fur
• + Highly expressive, larger-than-life eyes that convey a wholesome vibe
• + Consistent color palette across the different animals

• − Missed one of the four requested animals (red fox kit)
• − The animals are sitting rather than 'playfully chasing and tumbling'
• − The butterflies look like glowing blobs rather than detailed insects

GPT Image 1 Mini

• + Successfully included all four requested animals (dog, cat, bunny, and fox)
• + Dynamic action pose matches the 'chasing and tumbling' request perfectly
• + Realistic lighting with visible god rays and dew sparkles in the grass

• − The fox kit has slightly unusual black paws that look a bit muddy
• − The bunny's ears are a bit small compared to typical baby bunnies

FLUX1.1 [pro]

• + Excellent vintage illustration style with detailed etching on the cloche.
• + Beautiful composition with a classic border and subtle background texture.
• + Captures the 'warm brown and cream tones' perfectly.

• − Failed the primary text prompt, spelling it 'Cafeé FRATILIAN'.
• − The steam effect is very thin and almost looks like a crack in the image.

GPT Image 1 Mini

• + Perfect text rendering for both 'Caffè Florian' and 'EST. 1720'.
• + Clean, minimalist vector aesthetic that is very readable.
• + Clear interpretation of the steam element.

• − Ignored the request for a 'light background', providing a black one instead.
• − The cloche dome is very basic compared to the artistic vintage style requested.

FLUX1.1 [pro]

• + Excellent NASA-inspired color palette.
• + Sophisticated layout that feels like a real modern poster.
• + High-quality vector aesthetic with clean lines.

• − Text is largely gibberish or misspelt.
• − The numerical sequence is disordered or non-linear.
• − The large central moon dominates and confuses the actual trajectory steps.

GPT Image 1 Mini

• + Perfect adherence to the 6 requested steps in the correct order.
• + Text is perfectly legible and correctly spelled.
• + Strong iconography that matches the specific prompts for each step.

• − The 'Translunar' trajectory graphic is messy and contains weird loops.
• − Composition is a bit crowded and simplistic compared to a professional poster.
• − The background is a slightly off-white/beige rather than the light gray requested.