OpenAI's state-of-the-art image generation model with arbitrary resolution up to 4K and strong instruction following
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
GPT Image 2
#3 of 62 in Text-to-Image
Recraft V4.1 Pro
#36 of 62 in Text-to-Image
Where the votes landed
GPT Image 2
0%
win rate
Ties
0%
Recraft V4.1 Pro
0%
win rate
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
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 2
- + Excellent adherence to lighting instructions with a clear shadow cast from the left.
- + Highly realistic glass textures and refractions through the cube.
- + Perfect spatial arrangement of all requested objects.
- − The sphere is slightly opaque/matte compared to the glass cube.
Recraft V4.1 Pro
- + Aesthetically pleasing blue glass sphere effect.
- + Good composition with a more cinematic, moody feel.
- − The book is levitating slightly above the glass cube rather than sitting on it.
- − The glass cube has inconsistent thickness on its walls.
- − The lighting is more front-facing than coming from the left as requested.
Verdict: GPT Image 2 followed the prompt's spatial and lighting instructions perfectly, producing a physics-defying but visually grounded image. Recraft V4.1 Pro struggled with the physics of the book sitting on the cube and failed to capture the specific direction of the light as accurately.
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 2
- + Excellent photorealistic skin texture and fine details in the subject's face and hands.
- + Highly coherent bicycle mechanics and tool kit detail.
- + Subtle and naturalistic light rain and wet pavement effects.
- − The motion blur on the passing cars is very minimal compared to the prompt request.
- − The framing feels a bit too 'perfect' despite the prompt asking for imperfect framing.
Recraft V4.1 Pro
- + Strong execution of motion blur from passing cars that adds a sense of urban movement.
- + Effective 'imperfect framing' with a film-like aesthetic and slight vignette.
- + Visible rain streaks that clearly communicate the weather condition.
- − The man's hands are anatomically distorted and lack definition.
- − The bicycle's front fork and wheel alignment are physically impossible.
- − The skin texture is less detailed and realistic compared to Model A.
Verdict: GPT Image 2 is preferred for its superior technical execution, particularly in the rendering of the man's face, hands, and the mechanical detail of the bicycle. While Recraft V4.1 Pro captured the 'candid' motion blur and imperfect framing better, it failed significantly on the anatomical and structural details of the subject and the bicycle.
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 2
- + Excellent skin texture with realistic freckles and subtle dirt
- + High-quality engraving detail on the pauldrons
- + Very natural facial features and lighting
- − Hair beads are very small and less distinct than requested
- − Armor engraving feels a bit more weathered/faded compared to Model B
Recraft V4.1 Pro
- + Exceptional detail on the leather straps and textured cloth layers
- + Stronger adherence to the 'beads' prompt with distinct, colored beads
- + Dramatic bokeh sparks and high-contrast torchlight
- − Facial skin texture appears a bit muddy rather than just scarred/dirty
- − The eyes look slightly artificial compared to Model A
Verdict: Both models followed the prompt closely, but Recraft V4.1 Pro provided superior detail in the specific textures of the 'leather straps and cloth underlayer' while also rendering much more distinct beads in the hair. GPT Image 2 has more realistic facial aesthetics and skin rendering, but Recraft V4.1 Pro captures the 'battle-worn' and ornate aspects of the armor with higher clarity and punchier lighting.
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 2
- + Exceptional text rendering with perfect spelling and logical item descriptions.
- + Professional graphic design layout with consistent branding and iconography.
- + High-quality, vibrant food photography that aligns perfectly with the category headers.
- − The layout is quite dense, which might push the boundaries of 'minimalist' compared to a more spacious design.
Recraft V4.1 Pro
- + Features a clean and airy minimalist aesthetic with plenty of negative space.
- + Correctly includes the requested vibrant accents and sections.
- − Several spelling errors in the item descriptions (e.g., 'marinateed', 'sraalmon').
- − The grid layout for photos is disorganized, with food images placed somewhat randomly relative to the text sections.
- − The 'Pizza' photo is placed in the top right corner near the 'Appetizers' header, creating cognitive friction.
Verdict: GPT Image 2 (Model A) is vastly superior, producing a production-ready menu with perfect typography, logical organization, and high-quality imagery. While Recraft V4.1 Pro (Model B) attempts a minimalist style, it fails on basic functional requirements like spelling and logical layout, resulting in a confusing user experience.
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 2
- + Excellent typography integrated perfectly with the fiery theme
- + High-quality textures on the patty, bun, and vegetables
- + Strong adherence to the 'exploded' and 'dynamic' prompt requirements
- − The composition is a bit crowded with all elements packed tightly
- − The amount of sauce splashes might be slightly overwhelming for some
Recraft V4.1 Pro
- + Clean and modern commercial layout with good use of negative space
- + Accurately rendered text for all three requested messages
- + Realistic lighting on the food components
- − The 'LIMITED TIME ONLY' text lacks the requested fiery, glowing effect
- − The food components look more static and less cohesive than Model A
- − The starburst for the price is a bit generic
Verdict: GPT Image 2 (Model A) is the superior ad for this prompt as it perfectly captures the 'dynamic' and 'fiery' aesthetic across both the food and the typography. While Recraft V4.1 Pro (Model B) has a clean layout and excellent text legibility, it fails to apply the requested fiery effect to all text elements and lacks the energy found in Model A's composition.
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 2
- + Excellent adherence to the 'chalk texture' requirement with very realistic dusting and line quality.
- + Perfect rendering of all requested text including the truncated item which it logically completed.
- + High-quality composition that feels like a real café setting.
- − The cursive in the header is a bit stiff compared to a natural hand.
Recraft V4.1 Pro
- + Natural and varied handwriting styles that feel authentic to a person writing on a board.
- + Good use of depth of field and perspective for the chalkboard.
- + Correctly followed the prompt text for the main items.
- − The 'chalk' texture looks a bit more like a digital paint brush than actual dry chalk on a board.
- − The year '2026' in the header is slightly cramped compared to the rest of the text.
Verdict: GPT Image 2 (Model A) provides an incredibly realistic interpretation of the prompt, specifically excelling at the chalk texture and the overall coherence of the menu layout. While Recraft V4.1 Pro (Model B) has very convincing handwriting shapes, GPT Image 2's attention to detail in the chalk dust and material properties of the board makes it the superior output.
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 2
- + Excellent adherence to the specific 'horse on top' spatial instruction
- + Highly detailed textures on the spacesuit and lunar surface
- − The leather straps and reins are structurally confusing
- − Anatomically awkward back legs for the horse
Recraft V4.1 Pro
- + Beautiful cinematic lighting and galactic background
- + Clear and sharp details on the astronaut and horse
- − Completely failed the negative constraint to put the horse on top
- − Commonly seen trope that lacks the requested surrealism
Verdict: GPT Image 2 followed the complex spatial instruction perfectly, delivering the requested 'horse on top' surrealism. Recraft V4.1 Pro produced a higher quality cinematic image but failed the primary prompt constraint by placing the astronaut on top of 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 2
- + Excellent fur texture rendering on the capybara
- + Authentic NYC-style driver cap design
- + Cinematic lighting with realistic depth of field in the cabin
- − The capybara's paws look more like fingers/humanoid hands than actual paws
Recraft V4.1 Pro
- + Clean, professional composition with clear storytelling
- + Accurate depiction of capybara paws on the steering wheel
- + Realistic capture of the bored passenger expression and phone lighting
- − The hat is a generic flat cap rather than a structured taxi driver cap
- − The interior looks slightly generic and less detailed than the first image
Verdict: GPT Image 2 provides a more immersive and high-fidelity interior with superior lighting and texture, although the paws look a bit unnatural. Recraft V4.1 Pro handles the anatomy of the capybara's paws better and manages the placement of the business woman perfectly, but the overall image feels slightly less photorealistic than its competitor.
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 2
- + Excellent integration of text into the vintage gothic aesthetic.
- + Highly detailed border featuring thorns, webs, and skulls that matches the dark parchment theme.
- + Intricate background elements including a gothic castle and an arched bridge reflecting the location name.
- − Small scroll banner text is slightly cramped compared to model B.
Recraft V4.1 Pro
- + Clean, legible typography for the event details at the bottom.
- + Strong central focus on the glowing jack-o-lantern.
- + Good use of atmospheric fog and lighting effects.
- − The border feels more like a modern digital frame rather than an integrated vintage parchment part.
- − The scroll banner looks like a simple 2D overlay rather than being part of the scene's texture.
Verdict: GPT Image 2 is the superior choice as it perfectly captures the 'vintage gothic' aesthetic requested, with elaborate textures and a background that subtly references the location 'The Arches'. Recraft V4.1 Pro is very clean and readable, but it lacks the depth, intricate detail, and cohesive weathered feel that makes the first image look like an authentic invitation.
Isometric Miniature Diorama Scenes
Text-to-Image“Create a clear, 45° top-down isometric miniature 3D cartoon scene of Japan's signature dish: sushi, with soft refined textures, realistic PBR materials, gentle lighting, on a small raised diorama base with minimal garnish and plate. Solid light blue background. At top-center: 'JAPAN' in large bold text, 'SUSHI' below it, small flag icon. Perfectly centered, ultra-clean, high-clarity, square format.”
AI Judge Analysis
GPT Image 2
- + Excellent 3D rendering with realistic PBR materials and textures
- + Perfect typography and layout of text elements
- + Rich, detailed miniature scene that fits the dioramic theme
- − Includes more garnish and background elements than the 'minimal' request
Recraft V4.1 Pro
- + Successfully uses a minimal diorama base as requested
- + Clean, soft lighting suitable for a cartoon scene
- + Adheres well to the solid light blue background requirement
- − Text alignment and flag placement are slightly off-center
- − The sushi textures and modeling are less polished compared to the competitor
- − Text is flat 2D while the scene is 3D, creating a stylistic mismatch
Verdict: GPT Image 2 is the superior choice because its rendering of the 3D diorama is significantly higher quality with better textures and lighting. While Recraft V4.1 Pro followed the 'minimal' instruction more closely, it lacked the premium PBR look and professional typography layout found in the GPT output.
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 2
- + Excellent character consistency with all four animals clearly visible and facing the camera.
- + Vibrant color palette and strong god rays that fit the sunrise atmosphere.
- + Very clean rendering of the butterfly wings and flower details.
- − The composition feels a bit static, with the animals posed more like a group photo than 'tumbling'.
- − The backlight is slightly overwhelming, losing some detail in the upper highlights.
Recraft V4.1 Pro
- + Dynamic composition that perfectly captures the sense of 'tumbling' and 'playing'.
- + Superb texture work on the fur and realistic sparkling dew on the clover leaves.
- + Excellent lighting balance, preserving detail in both highlights and shadows.
- − The fox kit has a slightly distorted front paw while jumping.
- − The golden retriever's hind legs and tail area in the background look a bit anatomically confused.
Verdict: Recraft V4.1 Pro wins by a narrow margin because it captures the active, playful 'tumbling' motion requested by the prompt much more effectively than the static posing of GPT Image 2. While both models followed the prompt regarding the specific animals and lighting, Recraft V4.1 Pro's rendering of dew, fur textures, and natural interaction feels more like a 8K masterpiece.
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 2
- + Perfect adherence to text-based details including the accent on 'Caffè' and the banner date.
- + Exquisite vintage aesthetic with beautiful shading, hatching, and a paper-like texture.
- + Excellent composition with a balanced frame and professional typography.
- − Slightly more ornate than 'minimalist', though it fits the vintage brief perfectly.
Recraft V4.1 Pro
- + Clean, minimalist vector style that is highly scalable.
- + Correct spelling and iconography including the cloche and steam.
- − Text rendering is inconsistent with the 'n' in Florian appearing clipped or poorly integrated.
- − The composition feels a bit disjointed with the banner placed awkwardly behind the cloche base.
- − Lacks the 'vintage' texture and warmth requested in the prompt.
Verdict: GPT Image 2 provides a much more cohesive and professional logo that captures the 'vintage' and 'classic' requirements through sophisticated textures and typography. Recraft V4.1 Pro produces a more modern minimalist vector, but it suffers from poor letter spacing in the text and a less balanced layout compared to the polished output of its competitor.
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 2
- + Excellent typography and legible labels for all components.
- + Comprehensive and detailed illustration style that remains professional.
- + Includes accurate supplementary information like the crew names and landing site.
- − The illustrations use slightly more than 'subtle gradients', leaning towards a semi-realistic vector style.
- − The Apollo 11 insignia depicts an eagle carrying a branch in its claws rather than the classic mission patch style.
Recraft V4.1 Pro
- + Perfectly captures the requested 'flat-vector' style with crisp, minimal lines.
- + Adheres strictly to the navy and white color palette.
- + Uses very clean, modern iconography for a minimalist look.
- − Contains a counting error, labeling both descent and landing as step 05.
- − Included 'Saturn Vicon' as text instead of interpreting it as 'Saturn V icon'.
- − The text for 'Tranquility' has a minor spelling error.
Verdict: GPT Image 2 is the superior infographic, providing a much more informative and visually engaging layout that accurately follows the requested steps without numbering errors. While Recraft V4.1 Pro captures the 'flat-vector' aesthetic more authentically, it fails on logic (repeated step numbers) and literalizes prompt instructions ('Saturn Vicon') rather than understanding the intent.
Explore each model
Recraft V4.1 Pro tier text-to-image model — supports up to 2048×2048 and ultra-wide formats for hero imagery, campaigns, and print-ready work