Distilled version of HiDream AI's 17B parameter text-to-image model
Settled by community votes across 12 shared challenges, with an AI judge weighing in on each.
HiDream I1 Fast
18.2
arena score
#50 of 62 in Text-to-Image
Skill signature
· Text-to-Image
HiDream I1 Full
13.7
arena score
#60 of 62 in Text-to-Image
Vote tally
Where the votes landed
HiDream I1 Fast
0%
win rate
Ties
0%
HiDream I1 Full
0%
win rate
Shared challenges
12
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent adherence to lighting direction with realistic highlights
- + High-quality textures on the red book and wooden table
- + The sphere scale feels more grounded relative to the cube size
- − The glass back panel is invisible, making it look like an open frame rather than a cube
HiDream I1 Full
- + Successfully renders the cube as a solid object with visible back glass panels
- + Includes more realistic 'small' scale for the sphere as requested in the prompt
- + Captures the interaction of light through glass very convincingly
- − The sphere is floating unnaturally in the center of the air
- − Reflection of the sphere appears on the side glass wall in a physically inconsistent way
Verdict: HiDream I1 Full is the winner because it accurately depicts a 'small' sphere and clearly renders a 3D glass cube with all sides visible, whereas HiDream I1 Fast renders a sphere that is too large and a cube that looks like an open wireframe. While HiDream I1 Full features a floating sphere, its composition and adherence to the spatial requirements of the prompt are superior.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent street photography aesthetic with realistic bokeh and highlights.
- + More convincing wet pavement texture and reflections.
- + Better integration of the man with the bicycle parts.
- − The man's right hand consists of strangely fused fingers.
- − The car in the background is sharp rather than having the requested motion blur.
HiDream I1 Full
- + Stronger facial texture and realistic skin age details.
- + Captures the 'imperfect framing' request with a more candid feel.
- − Severe anatomical issues where the bicycle frame intersects through the man's leg.
- − The bicycle structure is nonsensical around the chain and seat post area.
- − Missing motion blur on cars.
Verdict: HiDream I1 Fast (Image A) produces a much more coherent scene, whereas HiDream I1 Full (Image B) suffers from major structural failures, notably the bicycle frame passing through the man's leg. While both models failed to generate the specific 'motion blur' requested for the cars, HiDream I1 Fast is the superior image due to its more believable composition and fewer glairing anatomical errors.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent engraving detail on the armor plate
- + Striking, high-contrast lighting with clear torch reflections
- + Cleaner facial rendering with lifelike eyes
- − The scars look a bit like face paint rather than physical wounds
- − Hair braids are a bit stiff in their positioning
HiDream I1 Full
- + Better 'battle-worn' aesthetic with more grit and blood on the face
- + Stronger bokeh effect with visible sparks in the background
- + Superior leather and strap texture detail across the chest
- − Slightly softer facial features compared to the armor
- − A strange dark artifact/smudge on the bridge of the nose
Verdict: Both models followed the prompt exceptionally well, producing high-quality cinematic portraits. HiDream I1 Fast delivers a cleaner, more heroic look with sharper detail on the metal engravings, while HiDream I1 Full leans more into the 'battle-worn' request with grittier skin textures and more realistic scarring. HiDream I1 Fast is slightly better overall due to the clarity of the facial features and the absence of the small visual artifact found on the nose in the Full version.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Clean grid layout for food photography
- + Strong use of bold sans-serif fonts in the header area
- − Nonsensical text in the main header
- − Garbled and illegible body text
- − Repetitive food photos that mainly show variations of the same pizza
HiDream I1 Full
- + Much more legible header and section text
- + Professional use of white space that fits the minimalist prompt
- + Food images are higher quality and integrated better with the text
- − Body text and descriptions are still gibberish
- − The layout is slightly vertical and less of a traditional 'grid' than Image A
Verdict: HiDream I1 Full is the superior model for this task because it generates legible headers and section titles that follow the prompt's instructions. While HiDream I1 Fast has a more complex grid, its header is illegible and the overall execution feels more cluttered and repetitive.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Successfully spelled most of the requested menu items like 'Truffle Mushroom' and 'Grilled Octopus'.
- + Captures a realistic café environment with depth of field.
- + Text has a convincing chalky texture with dusty artifacts.
- − Text displays severe overlapping and glitches in the middle section.
- − Spelling errors present like 'Buter' and 'Her's'.
HiDream I1 Full
- + Excellent layout with very clear, legible lettering.
- + High visual quality on the chalk illustration of the coffee cup.
- + Very clean presentation without overlapping text characters.
- − Completely failed to follow the specific menu item text requested in the prompt.
- − The text looks more like a digital font than natural hand-drawn chalk handwriting.
- − The content consists mostly of gibberish words and repetitive prices.
Verdict: HiDream I1 Fast attempted to follow the specific text instructions, resulting in recognizable menu items but suffered from significant rendering glitches and poor text alignment. HiDream I1 Full produced a much cleaner and more aesthetically pleasing image, but it completely ignored the content of the prompt, filling the board with nonsensical text. HiDream I1 Fast is the narrow winner for actually attempting the specific prompt requirements despite the technical artifacts.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Clean, photographic realism with high-quality textures
- + Correct anatomical rendering of the horse and astronaut gear
- − Failed the main spatial constraint, placing the character in a desert instead of space
- − Failed the surreal role-reversal prompt
HiDream I1 Full
- + Successfully captured the 'space' setting with a cinematic planetary backdrop
- + Dramatically improved prompt adherence regarding location
- − Failed the specific role-reversal instruction ('horse on top, not vice versa')
- − The horse anatomy is slightly distorted in the legs and hooves
Verdict: Both models failed the specific logic puzzle of the prompt ('horse on top, not vice versa'), which requested a surreal reversal of roles. However, HiDream I1 Full followed the 'space' and 'cinematic' instructions much better than HiDream I1 Fast, which produced a mundane desert scene.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent high-resolution fur texture on the capybara.
- + The blurred bokeh background captures the NYC night atmosphere perfectly.
- + Clear text and branding on the taxi driver cap.
- − The driver has human hands on the steering wheel instead of capybara paws.
- − The businesswoman is sitting in the front passenger seat instead of the back seat.
- − The capybara's head is not anatomically fused correctly with the body.
HiDream I1 Full
- + Successfully rendered capybara paws on the steering wheel.
- + The passenger is correctly placed in the back seat as requested.
- + Good lighting and shadows within the car interior.
- − The passenger's hands and phone appear distorted with multiple artifacts.
- − The capybara's head has a slightly detached appearance from its torso.
- − The passenger is sitting on the wrong side relative to typical taxi exits.
Verdict: HiDream I1 Full is the winner because it followed the spatial requirements of the prompt, correctly placing the passenger in the back seat and giving the capybara animal paws rather than human hands. While HiDream I1 Fast had better textures and background bokeh, the presence of human hands on the animal and the incorrect seating arrangement made it less accurate to the prompt.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Successfully rendered the numeric date 30.10.2026 correctly.
- + Captures the gothic aesthetic well with a vibrant glowing jack-o-lantern.
- + Includes the requested 'The Arches' location, even if garbled.
- − Text in the scroll banner is heavily garbled and incomplete.
- − Lower text area includes significant gibberish and overlapping characters.
- − The parchment edges feel slightly clipped by the square frame.
HiDream I1 Full
- + The scroll banner text 'You are invited to a night of frights' is perfectly legible.
- + The overall layout and composition are cleaner and more balanced.
- + The jack-o-lantern rendering is very sharp with excellent lighting.
- − Failed significantly on the specific date and time details, providing gibberish like 'Yootmber'.
- − Did not include the 'The Arches, NYC' location text requested.
- − Missed the word 'Invitation' in the main header title.
Verdict: HiDream I1 Fast followed the specific data requirements much closer by getting the date and location partially correct, whereas HiDream I1 Full failed on all specific event details. However, HiDream I1 Full produced a much cleaner visual composition and perfect banner text. HiDream I1 Fast is preferred for this specific task because the invitation actually contains the correct date, making it more functional despite the messier rendering.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent text legibility and graphic design integration
- + Accurate inclusion of the flag icon and additional requested elements like chopsticks
- + High clarity with a clean, professional aesthetic
- − Anatomically strange nigiri with a filling inside the rice block
- − Text contains some minor extraneous artifacts next to the word SUSHI
HiDream I1 Full
- + Beautiful 3D modeling with softer, more appealing textures
- + More complex and visually interesting arrangement of sushi
- + Cleaner background and better use of isometric space
- − Failed to include the requested flag icon
- − Text font is a bit more generic compared to model A
- − Minor color fringing on the left edge of the frame
Verdict: HiDream I1 Fast (Model A) followed the prompt more precisely by including the specific flag icon and creating a very clean graphic-style layout, though the sushi itself has some structural errors. HiDream I1 Full (Model B) produced a more aesthetically pleasing 3D render with superior lighting and textures, making the food look more appetizing, despite missing the flag icon. Model A is the technical winner for prompt adherence, while Model B is visually more impressive as a 3D isometric diorama.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Excellent fur texture rendering on the puppy and fox.
- + Subtle and naturalistic lighting within the meadow scene.
- + Good focus and depth of field.
- − Failed to include the requested baby bunny.
- − Included two kittens instead of one.
- − The animals are sitting still rather than 'playfully chasing butterflies and tumbling'.
HiDream I1 Full
- + Bright, vibrant colors that enhance the 'wholesome vibe'.
- + Clearer butterfly silhouettes.
- + Charming, expressive eyes on the animals.
- − Failed to include the requested baby bunny.
- − Has several anatomical artifacts, particularly the puppy's disjointed front legs.
- − Included two kittens instead of one.
Verdict: Both models failed to include the baby bunny or depict the requested tumbling/chasing action, opting instead for a static posed group shot. HiDream I1 Fast is the winner due to significantly better anatomical coherence and more realistic fur textures, whereas HiDream I1 Full suffers from a glaring error in the anatomy of the puppy's legs.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Successfully includes the primary business name 'Caffè Florian'
- + Excellent vintage aesthetic with appealing shading and stars
- + Dynamic and stylish interpretation of the steam elements
- − Typos in the text with '17210' instead of '1720'
- − The cloche dome is rendered as transparent or has confusing internal layering
HiDream I1 Full
- + Clean, minimalist vector style that adheres well to 'minimalist' prompt
- + Perfectly accurate text for 'Est. 1720'
- + Centered and balanced composition suitable for a logo
- − Completely missed the primary business name 'Caffè Florian'
- − The flat brown color is less visually interesting than the requested warm tones and textures
Verdict: HiDream I1 Fast (Model A) followed the brand name instructions and created a more visually engaging vintage design, but suffered from a typo in the date. HiDream I1 Full (Model B) achieved a much cleaner minimalist look and correct date, but failed the prompt's main requirement to include the name 'Caffè Florian'. Model A is the likely winner for successfully identifying the primary subject and capturing the 'vintage' spirit more effectively.
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.”
HiDream I1 Fast
HiDream I1 Full
AI Judge Analysis
HiDream I1 Fast
- + Attempts to follow the sequence of all six requested steps
- + Closer adherence to the NASA-inspired color palette with more gray and navy used correctly
- + Text is somewhat legible in places like 'LAUNCH' and 'DESCENG'
- − Graphic layout is messy with erratic lines and disjointed icons
- − Spelling errors are frequent and distracting
- − Final icon depicting the landing sequence is poorly rendered
HiDream I1 Full
- + Stronger vector graphic aesthetic with cleaner lines and professional composition
- + Excellent NASA-inspired header typography and logo integration
- + Visual elements are high-contrast and more balanced as a poster
- − Failed to include all six requested steps, stopping after step three
- − Included prompt instructions like '+ ICON' and '+ UNG' literally in the labels
- − The 'Earth Orbit' label is attached to a generic planet that looks more like Saturn
Verdict: HiDream I1 Fast attempted the full complexity of the prompt's six-step sequence but failed in terms of visual polish and clean lines. HiDream I1 Full produced a much higher quality vector illustration with superior layout and typography, but it significantly failed on prompt adherence by missing half of the requested infographic steps and literalizing prompt text in the labels. HiDream I1 Fast is the likely winner for better following the logical structure of the request despite lower technical quality.
Next steps
Explore each model
HiDream AI's 17B parameter text-to-image model using sparse diffusion transformer with mixture of experts, achieving state-of-the-art image generation quality with strong prompt following