Recraft's latest text-to-image generation model with high-quality output, supporting various aspect ratios and custom color palettes
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
Recraft V4
#11 of 62 in Text-to-Image
Stable Diffusion 3.5 Large
#29 of 62 in Text-to-Image
Where the votes landed
Recraft V4
85.7%
win rate
Ties
0.0%
Stable Diffusion 3.5 Large
14.3%
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
Recraft V4
- + Perfect adherence to spatial instructions including the book on top and sphere inside.
- + Highly realistic textures for the wood, paper, and glass.
- + Sophisticated lighting and reflections consistent with a window source.
- − The sphere appears to be floating without a visible support structure, which might look slightly unnatural.
Stable Diffusion 3.5 Large
- + Clean, modern aesthetic with bright lighting.
- + Accurately places a green plant behind the glass structure.
- − Fails the spatial prompt by putting the cube on top of the book and the sphere on top of the book.
- − The sphere is technically outside the glass cube (it is sitting on the book which is under the cube).
- − Harsh lighting creates slightly blown-out highlights on the wood.
Verdict: Recraft V4 followed all spatial instructions perfectly, placing the sphere inside the cube and the book on top. Stable Diffusion 3.5 Large struggled with the complex prepositional logic, reversing the order of the objects. Recraft V4 also produced a much more photorealistic image with better material rendering.
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
Recraft V4
- + Excellent handling of motion blur on passing cars as requested.
- + Authentic street photography feel with 'imperfect' framing and natural-looking background clutter.
- + Realistic rendering of wet pavement and light reflection.
- − The structural anatomy of the bicycle is broken, especially the front fork and wheel attachment.
- − The man's hand is poorly defined and blends into the bike frame.
Stable Diffusion 3.5 Large
- + Stronger anatomical realism of the subject's face and skin texture.
- + The bicycle structure is more coherent and recognizable.
- + Vibrant color palette and good use of shallow depth of field.
- − Failed to include 'motion blur' on the car in the background, which is sharp.
- − The lighting on the man's hair and clothes feels slightly like a studio setup rather than natural street lighting.
Verdict: Recraft V4 captures the 'candid street photo' aesthetic and complex prompt requirements like motion blur much better than the competitor. However, Stable Diffusion 3.5 Large produces a much cleaner and more anatomically correct primary subject, even if it ignores the request for motion blur on passing vehicles.
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
Recraft V4
- + Excellent depiction of multiple braids with beads as requested.
- + Realistic weathering with dirt and grit that feels integrated into the skin.
- + Outstanding detail on the leather straps and garment underlayers.
- − The sparks look like simple white dots rather than glowing embers.
- − The lighting is slightly flat despite the presence of the torch in the background.
Stable Diffusion 3.5 Large
- + Beautifully intricate engraving on the plate armor.
- + Complex and realistic hair texture and braiding.
- + Stronger contrast and more cinematic lighting composition.
- − Missed the request for beads in the hair braids.
- − The chainmail/underlayer texture is slightly mushy in certain areas compared to the armor.
Verdict: Recraft V4 followed the specific technical details of the prompt more closely, particularly regarding the beads in the hair and the texture of the leather straps. Stable Diffusion 3.5 Large produced a more aesthetically striking image with superior armor engraving, but it missed the small beads and the skin weathering was slightly less pronounced than in Recraft V4.
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
Recraft V4
- + Exceptional text rendering with perfect spelling of all menu items and headers.
- + Clean, professional grid layout that mimics a real-world digital or print menu.
- + Consistent lighting and high-quality product photography for all food items.
- − Minimalist aesthetic might feel a bit sparse to some users.
- − Some items are missing descriptions compared to others in the grid.
Stable Diffusion 3.5 Large
- + Vibrant color palette with a dynamic use of food photography borders.
- + Creative layout that feels more like a sophisticated brand identity.
- − Poor text rendering with numerous spelling errors like 'APPETIZRS' and 'MAIMAES'.
- − The grid layout feels cluttered and cuts off many of the food photos.
- − Fails to clearly separate the sections requested in an legible way.
Verdict: Recraft V4 is the clear winner as it produces a functional, professional menu with flawless text and a highly organized layout that perfectly matches the 'modern minimalist' prompt. Stable Diffusion 3.5 Large creates a more artistic composition but fails significantly on text legibility and practical usability, making it look like an AI-generated approximation rather than a usable 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
Recraft V4
- + Perfect adherence to text requirements, rendering all three phrases correctly.
- + Excellent interpretation of the 'exploded' layout with realistic ingredient physics.
- + Professional graphic design composition suitable for a real advertisement.
- − The lighting on the burger pieces is a bit flat compared to the fiery background.
Stable Diffusion 3.5 Large
- + Vibrant, cinematic lighting with high-quality fire and ember effects.
- + Very appetizing texture on the beef patties and melting cheese.
- − Failed to include any of the requested text.
- − Ignored the 'exploded' instruction, showing a stacked burger instead of suspended components.
- − Unconnected droplets and fire artifacts appearing randomly around the bun.
Verdict: Recraft V4 followed every instruction perfectly, including complex text rendering and a specific 'exploded' layout, resulting in a ready-to-use advertisement. In contrast, Stable Diffusion 3.5 Large completely ignored the text requirements and the specific structural request for an exploded view, providing a standard (though visually striking) burger image.
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
Recraft V4
- + Excellent text accuracy and legibility, including the truncated final item.
- + Realistic chalk texture and handwritten variations.
- + Good composition with relevant chalk-drawn illustrations.
- − The font choice, while looking handwritten, is slightly too uniform across different items.
- − Missed the request for 'elegant cursive' for the title.
Stable Diffusion 3.5 Large
- + Atmospheric cafe background with good lighting.
- + Good attempt at a complex chalkboard layout with framing.
- − Poor text rendering with numerous spelling errors ('TODAAY', 'Ottpuys', 'Cholcalte').
- − Incorrect date (2024 instead of 2026).
- − The font looks more like a digital chalk overlay than authentic handwriting.
Verdict: Recraft V4 is the clear winner as it perfectly rendered the complex menu text and adhered to the specific date requested. Stable Diffusion 3.5 Large struggled significantly with both spelling and following the textual instructions, resulting in a board filled with gibberish.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
Recraft V4
- + Excellent high-contrast cinematic lighting and textures
- + Highly detailed rock surfaces and horse anatomy
- + The composition creates a strong sense of depth
- − Fails the specific negative constraint; the astronaut is riding the horse
Stable Diffusion 3.5 Large
- + Beautiful dreamlike colors and nebular effects
- + Solid technical execution of the character and environment
- − Fails the specific negative constraint; the astronaut is riding the horse
- − The harness and reins have minor logical inconsistencies where they attach to the astronaut
Verdict: Both Recraft V4 and Stable Diffusion 3.5 Large failed the primary logical challenge of the prompt ('horse on top, not vice versa'), which was a test of spatial relationship understanding. Recraft V4 is the visually superior image due to its incredible cinematic lighting and sharp details, whereas Stable Diffusion 3.5 Large is more generic in its composition.
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
Recraft V4
- + Excellent adherence to the prompt including the businesswoman in the back seat.
- + Very realistic lighting and atmosphere that captures the feeling of a rainy NYC night.
- + High quality text rendering on the taxi cap.
- − The capybara's paws are rendered a bit strangely when gripping the wheel.
Stable Diffusion 3.5 Large
- + Sharp, high-contrast imagery with punchy colors.
- + Good detail on the capybara's fur and whiskers.
- − Completely failed to include the requested passenger in the back seat.
- − The capybara's left arm is anatomically awkward and detached from the body.
- − Lighting feels more like a studio than a realistic street scene.
Verdict: Recraft V4 is the clear winner as it followed every part of the prompt, including the specific requirement of a bored businesswoman in the passenger seat which Stable Diffusion 3.5 Large omitted entirely. Recraft V4 also achieved a much more convincing 'cinematic/photorealistic' look suitable for a rainy night in New York.
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
Recraft V4
- + Perfect text rendering of all requested event details.
- + Excellent atmospheric lighting and cohesive composition.
- + Sophisticated use of negative space for the text while maintaining spiderweb details.
- − The 'small scroll banner' is a simple white rectangle rather than a stylized scroll.
Stable Diffusion 3.5 Large
- + Creative use of a parchment paper texture as the background.
- + Good execution of the 'scroll banner' element with a vintage hand-drawn look.
- + Effective use of twisted trees and a large moon for mood.
- − Failed to include the specific event details (Date, Time, Location).
- − Text rendering on the banner is shaky with artifacts.
- − Lower overall image quality compared to Model A.
Verdict: Recraft V4 is the clear winner as it successfully rendered every piece of text requested in the prompt with high legibility and professional graphic design. While Stable Diffusion 3.5 Large captured the 'parchment' and 'scroll' vibes well, it failed to include the critical event details and had significantly worse image clarity.
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
Recraft V4
- + Excellent typography and graphic design integration.
- + High-realism materials for the sushi and ice base.
- + Clean, professional composition following all layout instructions.
- − Leaner more towards photorealism than the requested 3D cartoon style.
- − The crushed ice texture is a bit messy compared to a clean diorama base.
Stable Diffusion 3.5 Large
- + Perfectly captures the 3D cartoon miniature/toy aesthetic.
- + Creative use of signage and flags to incorporate text into the scene.
- + Strong rendering of soft, stylized PBR materials.
- − Failed to place the text at the top-center of the image as specified.
- − Composition is a bit cluttered with more garnish than requested.
Verdict: Recraft V4 followed the layout and typography instructions perfectly, placing the text and flag exactly where requested with high clarity, though it opted for a more realistic style. Stable Diffusion 3.5 Large better captured the '3D cartoon miniature' aesthetic but moved the required text into the scene elements and failed the specific layout placement.
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
Recraft V4
- + Excellent fur texture and realism in all four animals
- + Effectively captures the specific lighting effects like god rays and dew sparkles
- + Distinctly renders the fox, kitten, puppy, and bunny as requested
- − The bunny's paws look slightly amorphous and floating
- − Almost too sharp, pushing the photorealism into a slightly uncanny digital composite feel
Stable Diffusion 3.5 Large
- + Beautiful composition and bokeh effect in the background
- + Great sense of movement and 'tumbling' joy among the animals
- + Soft, warm lighting that perfectly matches the 'wholesome' vibe
- − The kitten and fox look very similar in facial structure
- − Lower overall detail in the fur texture compared to Model A
- − The butterfly on the rabbit's ear is poorly integrated
Verdict: Recraft V4 wins on technical execution and prompt adherence, providing high-detail textures and capturing all four specific animals with clear distinctions. Stable Diffusion 3.5 Large offers more artistic charm and better motion, but the animals look somewhat generic and the image lacks the '8K masterpiece' clarity of its competitor.
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
Recraft V4
- + Excellent typography with a sophisticated, custom feel
- + Proper spelling of 'Caffè' including the accent
- + Clean vector aesthetic with professional cross-hatching details
Stable Diffusion 3.5 Large
- + Successfully included the 'banner' and 'texture' requested in the prompt
- + Good use of the 'Est. 1720' as a centered element
- + Warm cream and brown tones perfectly match the vintage brief
- − Misspelled the primary name as 'Cafféé'
- − The cloche icon looks disorganized with clashing steam/fire elements
- − The banner construction is slightly asymmetrical
Verdict: Recraft V4 produced a much more professional logo with superior typography and correct spelling, though it missed the specific 'banner' element and texture requested. Stable Diffusion 3.5 Large followed the layout instructions more literally by including the banner and background texture, but it failed on basic spelling and icon clarity. Recraft V4 is the preferred choice for a real-world logo application due to its polish and design coherence.
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
Recraft V4
- + Perfectly followed the 6-step sequence with accurate text labels.
- + Excellent text rendering and clean, modern vector iconography.
- + Correctly adhered to the specified NASA-inspired color palette.
- − The trajectory line for the descent step is a bit simplistic compared to the rest of the art.
Stable Diffusion 3.5 Large
- + Detailed illustration style that captures an older NASA technical poster aesthetic.
- + Good use of the navy and muted red color palette.
- − Failed to follow the requested 6-step sequence, providing jumbled elements instead.
- − Unreadable and garbled text throughout the infographic.
- − Incorrectly included a Space Shuttle-style vehicle instead of a Saturn V rocket.
Verdict: Recraft V4 produced a professional-grade infographic that followed every instruction, specifically accurately numbering the steps and rendering all text perfectly. Stable Diffusion 3.5 Large failed on prompt adherence, including nonsensical text and incorrect spacecraft (a shuttle) while ignoring the requested logical flow.
Explore each model
Stability AI's 8.1-billion parameter Multimodal Diffusion Transformer (MMDiT) text-to-image model featuring improved image quality, typography, complex prompt understanding, and resource-efficiency