GPT Image 1.5 vs Recraft V4
Head-to-head across 12 challenges
GPT Image 1.5
50.0%
win rate
Ties
0.0%
Recraft V4
50.0%
win rate
Challenge Results
Geometric Composition
Text-to-Image“A glass cube on a wooden table. Inside the cube is a small blue sphere. On top of the cube sits a red book. A green plant is behind the cube, partially visible through the glass. Soft window light from the left.”
AI Judge Analysis
GPT Image 1.5
- + Excellent photographic realism and lighting
- + Physics-based interaction with the sphere sitting realistically on the bottom face
- + High-quality texture on the red book and wooden table
- − The glass cube has an unusually thick mirrored or double-layered bottom
Recraft V4
- + Very realistic window lighting and background depth
- + Accurate representation of a weathered, vintage red book
- + Strong composition and clear glass refraction
- − The blue sphere is floating unnaturally in the center of the cube
- − The cube lacks a visible top surface for the book to rest on, making the book appear to hover
Verdict: GPT Image 1.5 follows the physics of the scene much better, placing the sphere on the floor of the cube and the book clearly on top of a solid surface. Recraft V4 produces a more artistically diffused light and a lovely background, but fails on the physical arrangement as the sphere is floating and the top of the cube seems to have disappeared under the book.
Candid Street Photography
Text-to-Image“A candid street photo of an elderly Japanese man repairing a red bicycle in light rain, reflections on wet pavement, shallow depth of field, 50mm lens, natural skin texture, imperfect framing, motion blur from passing cars, cinematic but realistic, no stylization.”
AI Judge Analysis
GPT Image 1.5
- + Excellent skin texture and realistic raincoat material details
- + Shallow depth of field accurately follows a 50mm look
- + Consistent reflections and wet asphalt textures
- − The car in the background lacks the requested motion blur
- − Composition is a bit too clean and centered for an 'imperfect framing' request
Recraft V4
- + Successfully captures the motion blur from passing cars
- + More dynamic street atmosphere with pedestrians and bokeh
- + Effective use of vertical composition to enhance the rainy street feel
- − Anatomy issues with the man's hands and the bicycle's fork structure
- − Skin texture looks slightly smoothed and less 'natural' than Model A
- − The umbrella in the background has some merging artifacts
Verdict: GPT Image 1.5 excels in photographic realism and fine skin detail but fails to incorporate the specific 'motion blur' requested in the prompt. Recraft V4 captures the lively, imperfect energy of a rainy street scene better and adheres to the motion blur requirement, though it suffers from typical AI artifacts in the bicycle's geometry and the man's hands.
Fantasy Warrior
Text-to-Image“Close portrait of a battle-worn paladin in ornate engraved plate armor, hair braided with small beads, faint scars and dirt on the skin, warm torchlight reflecting off metal, shallow depth of field, bokeh sparks, lifelike eyes, highly detailed texture on leather straps and cloth underlayer.”
AI Judge Analysis
GPT Image 1.5
- + Excellent skin texture with realistic pores, scars, and dynamic reflections.
- + Superior composition with a tight close-up that emphasizes the armor's intricate engravings.
- + Warm, atmospheric lighting that perfectly captures the glow of torchlight on metal.
- − The bokeh background is slightly more generic and less defined than in Model B.
Recraft V4
- + Stronger adherence to the 'braided hair with small beads' detail across more of the hair.
- + Good depiction of the underlayer cloth and leather strap textures.
- + Atmospheric background including visible torches.
- − Skin texture appears somewhat muddy and lacks the lifelike detail of Model A.
- − Lighting feels flatter and less integrated into the environment than Model A.
- − Facial expression and features lack the sharpness and cinematic quality of the first image.
Verdict: GPT Image 1.5 is the clear winner due to its exceptional lifelike detail, specifically in the skin texture and the warm, reflective quality of the lighting on the armor. While Recraft V4 captured the braid and bead details more literally, the overall visual fidelity and cinematic composition of GPT Image 1.5 make it a much more compelling interpretation of the prompt.
Modern Clean Menu
Text-to-Image“Modern minimalist restaurant menu design, white background with colorful food photos in grid, sections for appetizers/pizza/mains, bold sans-serif fonts, vibrant accents, clean professional layout for casual dining.”
AI Judge Analysis
GPT Image 1.5
- + Excellent photographic quality with realistic lighting and texture.
- + Perfect text rendering with clear hierarchy and coherent pricing.
- + Highly functional and professional layout that looks like a real-world menu.
- − The 'Mains' section is cut off at the bottom.
- − Some food photos don't perfectly align with the text descriptions on the left.
Recraft V4
- + Strict adherence to the 'grid' prompt for food items.
- + Consistent visual style for all food photography with shadows on white backgrounds.
- + Extremely clean, minimalist aesthetic with good use of negative space.
- − Typos in text, such as 'solmon' instead of 'salmon'.
- − Slightly less 'vibrant' as requested, opting for a more sterile look.
- − Layout feels a bit repetitive and less like a finished professional print design.
Verdict: GPT Image 1.5 produces a more realistic and professional-looking menu with superior text rendering and high-quality food photography. While Recraft V4 followed the 'grid' instruction more literally, it suffered from spelling errors and a more clinical presentation. GPT Image 1.5 is the preferred choice for a usable, aesthetically pleasing design.
Magic Burger Explosion: Fiery Photorealism Challenge
Text-to-Image“Ad for 'Magic Burger'. Dynamic, exploded burger with all components (bun, patty, cheese, lettuce, tomato, sauce) suspended in mid-air. Emphasize photorealistic detail and a sense of motion. Dark, fiery background with glowing embers. Integrate text: 'MAGIC BURGER' as a prominent title, 'LIMITED TIME ONLY' as a secondary message, and '€6.99' in a starburst, all rendered with a fiery, glowing effect.”
AI Judge Analysis
GPT Image 1.5
- + Excellent adherence to the fiery, glowing text requirements
- + Highly detailed food textures with realistic lighting
- + Very dynamic composition with a strong sense of internal light and motion
- − The '€' symbol and some numbers have slightly messy edges
- − Small artifacts in the debris and floating sauce
Recraft V4
- + Natural and realistic food photography style
- + Clean and legible font for the secondary messaging
- + Good spatial separation of the exploded ingredients
- − The price is not inside a starburst as requested
- − The background and text effects lack the requested 'fiery, glowing' intensity
- − Missing the dramatic 'Magic' branding feel compared to the competitor
Verdict: GPT Image 1.5 performed significantly better on prompt adherence, capturing the specific aesthetic requests for fiery glowing text and a starburst price tag. While Recraft V4 produced very clean and realistic food assets, it failed to deliver the energetic 'Magic' atmosphere and specific design elements requested in the prompt.
Chalkboard Menu
Text-to-Image“Handwritten-style chalkboard menu in a cozy café, all text rendered in the exact same realistic chalk handwriting style with natural variations in letter size, slight slant, and chalk texture — no printed or digital fonts anywhere on the board. Title at the top in elegant cursive chalk handwriting: ‘TODAY’S SPECIALS – APRIL 30, 2026’. Below it, three menu items also in the same handwritten chalk style: ‘Truffle Mushroom Risotto – $24’, ‘Grilled Octopus with Lemon & Herbs – $28’, ‘Brown Butter Chocolate Chip Cookies – $9’. At the very bottom, smaller text in the identical handwritten chalk style (slightly smaller but still clearly legible with the same handwriting characteristics): ‘All items made fresh daily • Ask about our gluten-free options’. Warm ambient café lighting, visible chalk dust and smudges, realistic handwriting imperfections, no clean printed text anywhere.”
AI Judge Analysis
GPT Image 1.5
- + Excellent chalk texture throughout the handwriting
- + Natural variations in slant and letter height that mimic a human hand
- + Flawless rendering of all text and prices as requested
- − Missing the café background context, focusing only on the board
- − Visible artifacting in the top-right corner lighting
Recraft V4
- + Includes a realistic café background which adds to the requested cozy atmosphere
- + Solid text rendering with high legibility
- + Good chalk-like smudging on the board itself
- − The handwriting style is a bit more rigid and font-like compared to Model A
- − The cursive title isn't as elegant as requested
Verdict: GPT Image 1.5 followed the stylistic handwriting instructions more closely, producing text that looks genuinely hand-drawn with wonderful chalk texture and natural flow. While Recraft V4 successfully rendered the text and provided a better overall scene composition with the café background, its handwriting felt slightly more mechanical and less 'cursive' in the title.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
GPT Image 1.5
- + Excellent high-contrast lighting and sharp, cinematic details.
- + Dynamic composition with a sense of motion and dust kicks.
- + Includes multiple intricate celestial elements like Saturn and a lunar lander.
- − Failed the negative constraint; the astronaut is riding the horse instead of the horse riding the astronaut.
- − Anatomical issues with the horse's rear right leg.
Recraft V4
- + Strong atmospheric depth with the use of large foreground asteroids.
- + Consistent color palette providing a moody, surreal feel.
- + Very realistic textures on the horse's coat and mane.
- − Failed the negative constraint; the human-riding-horse trope was not inverted as requested.
- − Lower overall compositional complexity compared to the other model.
Verdict: Both models failed the specific instruction for the horse to be on top ('horse riding astronaut'). GPT Image 1.5 provides a much more detailed and vibrant scene with a rich background of planets and lunar equipment, whereas Recraft V4 offers a simpler, more monochromatic composition with better structural anatomy on the horse.
The Capybara Taxi Driver
Text-to-Image“Photorealistic scene inside a yellow New York taxi at night. A capybara is driving, wearing a yellow taxi driver cap and a dark jacket. It has a calm, professional expression and both front paws on the steering wheel. In the back seat sits a human businesswoman in a coat, looking at her phone with a completely normal, bored expression (as if this is just another normal ride). Through the windows you can see the streets of Manhattan at night with blurred lights. Realistic taxi interior, photorealistic, detailed fur and fabric, 35mm lens, night lighting with reflections, shallow depth of field.”
AI Judge Analysis
GPT Image 1.5
- + Excellent photorealism with shallow depth of field.
- + Perfect character expression and positioning of the paws on the wheel.
- + Accurate and high-quality rendering of the taxi driver cap.
- − The composition is a bit tight, showing less of the car's interior.
Recraft V4
- + Wider composition captures more of the environment and the business woman.
- + Excellent integration of textures like the leather seats and checkered coat.
- + Good interpretation of the bored expression requested for the passenger.
- − The capybara's head is disproportionately large and looks somewhat like a mask.
- − The driver's paws do not interact realistically with the steering wheel.
Verdict: GPT Image 1.5 wins on photographic quality and character realism, making the absurd concept feel truly grounded in reality. While Recraft V4 offers a better wide-angle composition and great environmental details, the distorted proportions of the capybara and its poorly rendered hands make it less convincing than the polished execution of GPT Image 1.5.
The Halloween Invitation
Text-to-Image“Vintage gothic Halloween party invitation. Dark parchment poster, spooky border with webs and thorns, central glowing jack-o-lantern, bats, twisted trees, moody night sky. Add elegant gothic title text saying "Halloween Party Invitation", a small scroll banner saying "You are invited to a night of frights", and event details at the bottom: Date: 30.10.2026 Time: 7pm Location: The Arches, NYC Spooky but polished, cinematic lighting, square format.”
AI Judge Analysis
GPT Image 1.5
- + Excellent adherence to the vintage parchment aesthetic with a warm, sepia-toned palette.
- + The typography is beautifully integrated into the layout with elegant gothic styling and decorative flourishes.
- + The border of thorns and webs is intricate and perfectly matches the prompt's description.
- − The text at the very bottom is slightly off-center compared to the main title.
- − The background elements like the graveyard and castle are a bit muddy.
Recraft V4
- + High clarity and sharp details on the pumpkin and the swarm of bats.
- + Clean, legible text for the event details at the bottom.
- + Creative use of rainy atmosphere to enhance the 'moody night sky' requirement.
- − The white scroll banner is too modern and clean, clashing with the 'vintage gothic' prompt.
- − The overall composition feels less like a cohesive 'invitation poster' and more like a scene with text overlaid.
- − Lacks the 'dark parchment' texture requested in the prompt.
Verdict: GPT Image 1.5 follows the stylistic requirements of the prompt much more effectively, delivering a cohesive vintage parchment aesthetic with superior typography and borders. While Recraft V4 has sharper individual elements and clean text, it fails to capture the 'vintage gothic' feel, using a banner that looks out of place and ignoring the parchment texture.
Adorable Baby Animals in Sunny Meadow
Text-to-Image“Hyper-photorealistic scene of fluffy baby animals—a golden retriever puppy, tabby kitten, baby bunny, and red fox kit—with big expressive eyes and ultra-detailed soft fur, playfully chasing butterflies and tumbling together in a lush wildflower meadow, warm golden sunrise light with god rays and dew sparkles, joyful wholesome vibe, 8K masterpiece.”
AI Judge Analysis
GPT Image 1.5
- + Excellent depiction of the 'tumbling together' prompt through dynamic, close-up interaction
- + Vibrant, warm lighting with strong god rays and dew sparkles
- + Expressive facial features that match the joyful, wholesome vibe
- − Anatomy on the kitten's paws is slightly messy
- − The fox loses its distinct body structure by being merged into the cuddle pile
Recraft V4
- + Stronger 'chasing butterflies' action with animals clearly in motion
- + Crisp fur textures and accurate anatomy for each of the four distinct species
- + Better distribution of elements across the composition including varied flora
- − The kitten's expression is a bit vacant compared to the others
- − Lighting feels slightly more synthetic and less 'warm golden' than requested
Verdict: GPT Image 1.5 captures the warm, emotional essence of the prompt through a cozy, tumbling composition and beautiful golden-hour lighting. Recraft V4 provides a more technically accurate and spacious scene that better represents the 'chasing' aspect of the prompt with incredibly clean fur details and balanced species representation.
Vintage Cafe Logo
Text-to-Image“Vintage minimalist restaurant logo for "Caffè Florian", retro cloche dome with steam and "Est. 1720" banner, classic typography, warm brown and cream tones, subtle texture on light background, vector emblem style.”
AI Judge Analysis
GPT Image 1.5
- + Excellent typography with a professional vintage flair
- + Beautiful use of texture and shading on the cloche dome
- + Perfectly executed banner for the establishment date
- − Failed to provide a light background as requested
- − Minor alignment issue with the steam centered over the dome handle
Recraft V4
- + Successfully followed the light background instruction
- + Clean vector style that feels modern yet minimalist
- + Accurate text rendering and placement
- − The 'Est. 1720' is in a semicircle rather than the requested banner
- − Steam looks a bit thin and less integrated compared to the rest of the logo
Verdict: GPT Image 1.5 produced a much more sophisticated and aesthetically pleasing logo with excellent vintage typography and texture, though it completely ignored the requirement for a light background. Recraft V4 followed all prompt instructions including the background color, but its design is simpler and missed the 'banner' element, substituting it with a basic shape. GPT Image 1.5 is the preferred choice for its superior artistic execution and higher-quality logo design.
Apollo 11: Journey to Tranquility
Text-to-Image“Create a clean, modern vector infographic poster about the Apollo 11 mission. NASA-inspired palette (navy, white, muted red, light gray). Flat-vector style, crisp lines, consistent iconography, subtle gradients only. Steps (stop at landing): 1. Launch (Saturn Vicon) 2. Earth Orbit (Earth + orbit ring icon) 3. Translunar (trajectory arc icon) 4. Lunar Orbit (Moon + orbit ring icon) 5. Descent (lunar module descending icon) 6. Landing (lunar module on the surface icon) Small supporting elements (minimal text): • Crew strip: three silhouette icons with only last names: Armstrong, Aldrin, Collins. • Landing site marker: Moon pin labeled "Tranquility" only. Layout constraints: generous margins, large readable labels, clean background with subtle stars. Vector-only, print-poster look, high resolution.”
AI Judge Analysis
GPT Image 1.5
- + Excellent typography with no spelling errors across many labels.
- + Logical and clear flow that follows the requested 6-step sequence perfectly.
- + Strong technical illustration style that creates a cohesive poster layout.
- − The 'Descent' icon looks slightly cluttered compared to the other minimalist steps.
Recraft V4
- + Elegant, minimalist vector aesthetic with excellent use of whitespace.
- + High-quality custom iconography for the crew and the lunar surface.
- + Crisp text rendering and consistent line weights.
- − The visual flow is a bit disjointed, with steps arranged sporadically rather than in a clear sequence.
- − Missed the 'Translunar' iconography requested (it is just a line with text, whereas Model A used a moon icon).
Verdict: GPT Image 1.5 is the winner because it provides a much more functional and readable infographic that follows the numbered steps in a logical narrative flow. While Recraft V4 has a very high-end minimalist design aesthetic, its layout is harder to follow as a 'step-by-step' guide, though both models handled the complex text requirements and color palette perfectly.
GPT Image 1.5
OpenAI's state-of-the-art image generation model with better instruction following and adherence to prompts
Recraft V4
Recraft's latest text-to-image generation model with high-quality output, supporting various aspect ratios and custom color palettes