Distilled version of HiDream AI's 17B parameter text-to-image model
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
HiDream I1 Fast
#49 of 62 in Text-to-Image
Wan 2.5 (Preview)
#27 of 62 in Text-to-Image
Where the votes landed
HiDream I1 Fast
0%
win rate
Ties
0%
Wan 2.5 (Preview)
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
HiDream I1 Fast
- + Excellent clarity on the glass cube edges
- + Natural-looking reflections on the blue sphere
- + Good adherence to the 'soft light' lighting description
- − The blue sphere is quite large, ignoring the 'small' descriptor in the prompt
- − The vertical edges of the cube appear slightly disconnected at the top right
Wan 2.5 (Preview)
- + Realistic texture on the red book with aged edges
- + High quality rendering of light rays and dust motes
- + Accurate glass reflections on the table surface
- − The blue sphere's reflection is floating and doesn't align with its base
- − The sphere appears to be levitating slightly rather than sitting on the bottom of the cube
Verdict: Both models followed the complex spatial instructions well, but HiDream I1 Fast produced a cleaner, more cohesive image. While Wan 2.5 (Preview) offered richer textures and more atmospheric lighting, it suffered from perspective issues where the sphere and its reflection did not align correctly with the cube's floor.
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
HiDream I1 Fast
- + Captures the requested motion blur on background cars
- + Excellent wet pavement reflections
- + Atmospheric shallow depth of field
- − Anatomy issues with the subject's feet and the bicycle frame
- − Bicycle is missing pedals and has an illogical seat/frame connection
Wan 2.5 (Preview)
- + Superior anatomy and realistic skin texture
- + Highly detailed bicycle and tools showing actual repair work
- + Excellent rendering of falling rain and rain splashes on the ground
- − Lacks the requested motion blur from passing cars
- − Framing feels a bit too perfect relative to the 'imperfect framing' prompt
Verdict: Wan 2.5 (Preview) produces a significantly more coherent and realistic image, with a believable subject and a detailed, functional-looking bicycle. While HiDream I1 Fast better captures the specific request for motion blur on the cars, its structural errors—such as the missing pedals and the man's feet merging with the bike—make it a less successful image overall.
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
HiDream I1 Fast
- + Excellent intricate engraving on the plate armor
- + Impressive 'battle-worn' detailing with visible facial scars
- + Rich lighting and bokeh effect in the background
- − Physical anatomy feels slightly AI-standardized and glossy
- − The hair and beads look a bit stiff and artificial
Wan 2.5 (Preview)
- + Exceptional lifelike skin texture and realistic, haunting eyes
- + Superior rendering of leather straps and frayed cloth underlayers
- + More naturalistic lighting and depth of field
- − Armor engravings are slightly less sharp than Model A
- − Character appears younger than a typical 'battle-worn' paladin archetype
Verdict: Wan 2.5 (Preview) wins due to its incredible realism in skin texture, fabric details, and lighting, which feels more cinematic and less 'digital' than its counterpart. While HiDream I1 Fast has excellent armor engraving and a strong fantasy aesthetic, it lacks the fine textural nuances seen in the leather and cloth transitions of Wan 2.5.
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
HiDream I1 Fast
- + Strong bold font in the header
- + Good use of high-contrast vibrant red accents
- − Text is largely illegible and gibberish
- − Food images are repetitive and lack diversity in categories
- − Layout feels cramped and poorly balanced
Wan 2.5 (Preview)
- + Excellent adherence to the grid layout request
- + Followed section instructions for appetizers, pizza, and mains perfectly
- + High visual quality and consistent aesthetic for food photos
- − Several spelling errors in the headings (e.g., 'Restormalit Menue')
- − Small body text is comprised of illegible character symbols
Verdict: Wan 2.5 (Preview) is the clear winner as it followed all layout and content instructions, including specific food categories and a clean grid design. HiDream I1 Fast failed to distinguish between sections and provided repetitive pizza images, whereas Wan 2.5 (Preview) produced a professional, minimalist aesthetic suitable for a real restaurant.
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
HiDream I1 Fast
- + Successfully integrated all requested text with clear readability.
- + The price starburst and primary title are prominent and well-arranged.
- + The fiery coals at the base create a strong thematic atmosphere.
- − Failed to create an 'exploded' burger view, keeping the ingredients mostly stacked.
- − The 'limited time' text contains spelling errors and overlapping characters.
- − Floating white onion rings appear somewhat detached and unnatural compared to the main burger.
Wan 2.5 (Preview)
- + Excellent 'exploded' view with dynamic suspension of all components.
- + High visual quality with sharp details on the grilled patty and lettuce.
- + Creative dripping-lava effect on the typography matches the fiery theme perfectly.
- − The price starburst is slightly less integrated into the overall composition than image A.
- − Missing the specific fiery ground/embers at the very bottom requested in the 'dark fiery background'.
Verdict: Wan 2.5 (Preview) significantly outperformed HiDream I1 Fast on the core 'exploded' burger concept, delivering a much more dynamic and professional-looking advertisement. While HiDream I1 Fast captured the fiery ground better, its failure to separate the burger layers and the spelling errors in the secondary text make it the weaker choice compared to the high-detail execution of Wan 2.5.
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
HiDream I1 Fast
- + Excellent photographic quality and café background depth.
- + The cursive handwriting for 'Today's Specials' is elegant and matches the prompt's style.
- − Numerous text artifacts and overlapping characters on the menu items.
- − Failed to correctly spell 'Butter' and 'Risotto'.
- − Visual layout of prices and text is disorganized and cluttered.
Wan 2.5 (Preview)
- + Near-perfect adherence to the text prompt and complex menu names.
- + Excellent chalk texture and natural variations in handwriting style.
- + Clean and professional layout that looks like an actual handwritten board.
- − The 'Today's Specials' title is more of a print-block style than the requested elegant cursive.
Verdict: Wan 2.5 (Preview) is the clear winner as it successfully rendered nearly all the complex text and pricing instructions with minimal errors and a very realistic chalk texture. In contrast, HiDream I1 Fast suffered from significant text hallucinations, misspellings, and jumbled characters despite having a nice background composition.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
HiDream I1 Fast
- + Natural-looking lighting and textures on the horse
- + Realistic composition of an equestrian setup
- − Completely failed the setting, placing the subject in a desert instead of space
- − Ignored the specific 'horse on top' instruction
- − Composition is very literal and lacks the requested surrealism
Wan 2.5 (Preview)
- + Correctly interpreted the 'in space' setting with cinematic lighting
- + High level of detail on the space suit and nebulae
- + Dynamic and vibrant composition
- − Failed the specific 'horse on top, not vice versa' logic flip requested in the prompt
- − Noticeable anatomical artifacts with the horse's legs merging together
Verdict: Both models failed the specific logical constraint of placing the horse on top of the astronaut, instead providing the common 'astronaut on horse' trope. However, Wan 2.5 (Preview) is the superior image as it actually followed the instruction to place the scene in space, whereas HiDream I1 Fast generated a standard desert environment.
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
HiDream I1 Fast
- + Features a very high-quality Capybara head with realistic fur texture.
- + Vibrant and high-contrast night lighting on the city background.
- − The driver's body consists of human hands and arms rather than capybara paws.
- − Failed to place the passenger in the back seat, placing her in the passenger seat instead.
Wan 2.5 (Preview)
- + Excellent adherence to the layout prompt with the capybara in the driver's seat and human in the back seat.
- + Accurately depicts animal-like paws on the steering wheel.
- + The capybara's expression and clothing perfectly match the 'professional' and 'dark jacket' description.
- − The text on the taxi meter/sign is gibberish.
- − The rainy texture on the windshield slightly obscures some of the interior detail.
Verdict: Wan 2.5 is the clear winner as it followed all the complex spatial instructions, correctly placing the human passenger in the back seat and ensuring the driver had paws instead of human hands. HiDream I1 Fast failed on both the passenger's position and the biological accuracy of the capybara driver, showing human hands instead of paws.
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
HiDream I1 Fast
- + Strong composition with a clear parchment focal point
- + Includes most elements like the jack-o-lantern and bats
- + Correctly identifies the date and location in the text
- − Significant text rendering issues and typos in the scroll and event details
- − Lacks the 'thorns' element mentioned in the prompt
- − Lower overall visual fidelity and more simplistic illustration style
Wan 2.5 (Preview)
- + Near-perfect text rendering for the title, scroll, and event details
- + Highly detailed and cinematic lighting on the jack-o-lantern and trees
- + Executes the 'webs and thorns' border perfectly
- − The parchment background is circular rather than a full poster sheet
- − Slight overlap of the bats and the text at the top
Verdict: Wan 2.5 (Preview) is the clear winner due to its superior text rendering and high-quality artistic execution of complex elements like the thorn border and cinematic lighting. While HiDream I1 Fast captures the general layout of a parchment poster, it fails on legibility and fine detail compared to Wan 2.5.
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
HiDream I1 Fast
- + Excellent text rendering and layout placement
- + Higher variety of sushi types on the diorama
- + Accurate 45-degree isometric perspective
- − Nigiri sushi contains an anatomically incorrect drawing inside the rice
- − Small artifacts near the 'SUSHI' text and flag icon
Wan 2.5 (Preview)
- + Superior PBR material rendering with realistic subsurface scattering
- + Cleaner, professional lighting and shadows
- + Ultra-clean text rendering with integrated flag
- − Lacks the requested 'small raised diorama base' (uses a simple round plate)
- − Less variety in the food shown compared to Model A
Verdict: Both models followed the prompt well, but HiDream I1 Fast better adhered to the 'isometric diorama' requirement with more complex sushi assets. However, Wan 2.5 (Preview) produced a much higher quality render with better lighting and material physics, though it simplified the scene to a single piece of sushi on a plate.
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
HiDream I1 Fast
- + Excellent fur texture rendering
- + Natural, soft lighting that blends the subjects well into the environment
- + Clean, high-resolution appearance with minimal artifacts
- − Failed to include the requested baby bunny
- − Includes two kittens instead of one, missing animal variety
- − The composition is static despite the prompt's request for tumbling and chasing
Wan 2.5 (Preview)
- + Successfully included all four requested animals (dog, cat, bunny, fox)
- + Captured the dynamic action of 'chasing' and 'tumbling' much better
- + Included specific details like god rays and floating dew sparkles
- − The fox's eyes appear unnaturally blue and slightly distorted
- − Some anatomical issues with the kitten's leg positioning
- − The dew drops appear as floating spheres rather than resting on surfaces
Verdict: Wan 2.5 (Preview) followed the prompt much more accurately by including all four distinct animal types and depicting a sense of motion. HiDream I1 Fast produced a cleaner image with better realism in lighting, but failed the prompt instructions by omitting the bunny and including extra kittens. While Wan 2.5 has some slight anatomical glitches, its adherence to the specific 'chasing' action and subject list makes it the superior response.
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
HiDream I1 Fast
- + Successfully included all prompt elements like cloche, steam, and banner.
- + Followed the warm brown and cream color scheme.
- − The date '17210' is a typo that does not follow the prompt.
- − The 'Caffè Florian' text placement inside the cloche is slightly cluttered.
- − The steam lines are somewhat awkward and detached from the dome.
Wan 2.5 (Preview)
- + Excellent typography and correct spelling of all text, including the date.
- + Superior vector emblem composition with a clean, professional layout.
- + The cloche and steam elements are integrated much more naturally.
- − The parchment texture is a bit more heavy-handed than the 'subtle' request.
Verdict: Wan 2.5 (Preview) is the clear winner as it produced a professional, perfectly spelled, and well-composed logo that adheres strictly to the prompt. HiDream I1 Fast failed on the text rendering, providing an incorrect date (17210) and a messy placement of the brand name inside the graphic element.
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
HiDream I1 Fast
- + Follows the requested color palette well with navy, white, and red.
- + Maintains a consistent, clean white background for the infographic area.
- − Text is poorly rendered with significant misspellings like 'DESCENG' and 'EOH OPBIT'.
- − Internal logic of the infographic is messy, with icons and labels disconnected or incorrectly sequenced.
- − Image quality is blurry with visible artifacts around the shapes.
Wan 2.5 (Preview)
- + Excellent typography with legible labels for Launch, Earth Orbit, Translunar, and Lunar Orbit.
- + High visual quality with a clear vector-style aesthetic and professional layout.
- + Creative addition of the Apollo 11 crew icons and names, plus the landing site marker.
- − Includes a space shuttle/generic rocket hybrid instead of a faithful Saturn V icon.
- − The 'Descent' and 'Landing' text labels lack their own specific icons, unlike the earlier steps.
Verdict: Wan 2.5 (Preview) is the clear winner as it produces a professional, legible, and logically structured infographic with correctly spelled labels and high-quality vector art. HiDream I1 Fast fails significantly on text rendering and structural coherence, resulting in a confusing layout with numerous spelling errors.
Explore each model
Alibaba's text-to-image and image-to-image generation model from the Wan AI suite, offering high-quality visual generation capabilities