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
Settled by community votes across 16 shared challenges, with an AI judge weighing in on each.
HiDream I1 Full
#60 of 62 in Text-to-Image
Z-Image Turbo
#12 of 62 in Text-to-Image
Where the votes landed
HiDream I1 Full
0%
win rate
Ties
0%
Z-Image Turbo
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 Full
- + Excellent refraction and reflection physics, showing the blue sphere and plant distorted through the glass edges.
- + Rich, warm lighting and detailed wood grain texture on the table.
- + Superior depth of field with the plant clearly visible through the glass as requested.
- − The blue sphere appears to be floating in the center rather than resting on the bottom.
- − The book texture is slightly less paper-like than the other model.
Z-Image Turbo
- + Very realistic book texture with visible wear on the edges.
- + Clean, modern composition with a clear focus on the subject.
- − Fails to show the green plant through the glass of the cube, as it is positioned too high and to the side.
- − The reflection of the sphere on the bottom of the cube looks like a mirror reflection rather than clear glass sitting on wood.
- − The glass cube lacks thickness and realistic edge refraction.
Verdict: HiDream I1 Full is the winner because it successfully captured the complex interplay of light and transparency, showing the plant through the glass as requested. While Z-Image Turbo produced a high-quality book, it failed the specific spatial requirement of having the plant visible through the cube and had less convincing glass physics.
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 Full
- + Excellent shallow depth of field and bokeh effects.
- + Beautiful reflections on the wet pavement that add to the cinematic atmosphere.
- + High level of detail in the subject's face and clothing texture.
- − The bicycle frame geometry is physically impossible and lacks a proper seat post connection.
- − The cars in the background look static rather than having the requested motion blur.
Z-Image Turbo
- + Natural skin textures and candid, realistic posture.
- + The red bicycle has more convincing mechanical components compared to the other model.
- + Visible rain streaks fall across the frame, enhancing the weather atmosphere.
- − Lacks the requested motion blur from passing cars.
- − Shadows and lighting feel a bit flat, missing the requested cinematic depth.
- − The composition is quite tight at the edges, losing some of the street context.
Verdict: HiDream I1 Full produces a much more cinematic and visually pleasing image with great use of depth of field and reflections, though it fails significantly on the anatomy of the bicycle. Z-Image Turbo provides a more grounded, realistic scene with better bicycle mechanics, but feels more like a standard snapshot than the cinematic request. HiDream I1 Full is the winner for its superior lighting and adherence to the 'cinematic' and 'shallow depth of field' requirements.
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 Full
- + Strong symmetrical composition and intense facial expression.
- + High-contrast lighting that emphasizes the engraved armor textures well.
- + Accurate representation of the requested beads in the hair braids.
- − The scars appear slightly artificial, looking more like surface paint than healed skin.
- − Evidence of AI artifacts in the hair, particularly how it blends into the leather strap.
Z-Image Turbo
- + Excellent realism in skin texture, dirt, and lifelike eyes.
- + Masterful lighting where the torchlight naturally interacts with the metal and face.
- + Superior rendering of complex textures like the chainmail and quilted cloth underlayer.
- − The compositions is slightly off-center compared to the 'close portrait' request.
- − The sparks from the torch are a bit sharp and distracting compared to the requested bokeh effect.
Verdict: Z-Image Turbo is the clear winner due to its superior realism and natural integration of textures, particularly in the facial details and sub-armor layers. While HiDream I1 Full offers a striking and symmetrical composition, Z-Image Turbo captures a more authentic 'battle-worn' aesthetic with more convincing lighting and material rendering.
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 Full
- + Crisp and clear high-resolution food photography
- + Consistent food style (all pizzas) creates a cohesive look
- + Excellent text legibility and clean sans-serif typography
- − Lack of food variety as requested by 'colorful food photos in grid'
- − Layout is a bit sparse with too much white space between elements
Z-Image Turbo
- + Successfully implements the 'food photos in grid' request with varied dishes
- + Professional use of vibrant orange accents as requested
- + More complex and realistic menu layout with multiple pricing tiers
- − Contains significant spelling errors like 'PIZZA MANS' and 'SE IIION'
- − Food images are slightly lower in clarity compared to Model A
Verdict: Z-Image Turbo followed the layout instructions much better by providing a variety of food items in a grid with the requested vibrant accents, though it struggled with text accuracy. HiDream I1 Full produced much higher quality individual images and cleaner text, but it failed to provide a diverse food grid, showing only pizzas in every section. Z-Image Turbo is the likely winner for better adhering to the complex design requirements of 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
HiDream I1 Full
- + Features a beautiful, high-quality wooden frame and cafe background.
- + The chalk illustration of the coffee cup is visually appealing and detailed.
- − Completely failed to render the requested menu items, using gibberish text instead.
- − The text style looks like a digital font rather than natural chalk handwriting.
Z-Image Turbo
- + Excellent prompt adherence, including the specific menu items and prices.
- + The chalk texture is highly realistic with believable smudges and hand-lettered variations.
- − Minor spelling error in 'Mustroom' for 'Mushroom'.
- − Composition is a bit tight with the wooden border barely visible on the sides.
Verdict: Z-Image Turbo is the clear winner as it followed the complex text instructions perfectly, rendering the specific menu items and dates requested. HiDream I1 Full failed on all specific text requirements, filling the board with nonsensical words and repeated price tags, despite having a more complete overall 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 Full
- + Strong cinematic composition with the Earth's curve and clouds in the background.
- + Good anatomical consistency for the horse and astronaut interaction.
- + Vibrant lighting and high level of detail on the horse's fur and the suit.
- − The horse's back hooves look slightly distorted and blurry.
Z-Image Turbo
- + Clear lighting on the astronaut and horse subjects.
- + The space background is minimalist and clean.
- − Failed the negative constraint; the prompt specifically asked for 'horse on top' (a horse riding an astronaut), but this shows an astronaut riding a horse.
- − The background is less 'cinematic' and lacks the depth seen in the competitor.
- − The horse's hind legs have awkward proportions and positioning.
Verdict: Both models failed to correctly interpret the specific surreal prompt instruction 'horse on top, not vice versa,' as both depicted an astronaut riding a horse. However, HiDream I1 Full is the superior image due to its much better cinematic composition, more detailed textures, and more realistic integration of the horse and rider within the space environment. Z-Image Turbo has anatomy issues with the horse's legs and a much flatter background.
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 Full
- + Excellent texture on the capybara's fur and the jacket
- + Clear and legible 'TAXI' text on the hat
- + Accurately depicts the bored expression of the businesswoman in the back seat
- − The capybara's hands look more like human-monkey hybrids than capybara paws
- − Composition makes/the businesswoman appear to be sitting much closer to the driver than in a typical taxi
Z-Image Turbo
- + Authentic chauffeur-style hat that adds to the 'professional' driver persona
- + Stronger sense of cinematic lighting and bokeh in the background
- + Good depth of field separating the driver from the passenger
- − The capybara's hands are fused awkwardly with the steering wheel
- − The passenger is looking away from her phone rather than at it as requested
- − Minor anatomical issues where the fur meets the jacket
Verdict: HiDream I1 Full is the winner because it adhered more closely to the prompt specifics, particularly the businesswoman's expression and her interaction with the phone. While Z-Image Turbo captures a nice atmosphere, HiDream I1 Full provides better text rendering on the hat and more detailed fur textures.
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 Full
- + Crisp, clean illustration style
- + Excellent use of the central banner for the invite text
- + Coherent color palette
- − Failed significantly on the specific date and time text
- − Omitted the location details entirely
- − The text 'Halloween Party Invitation' is abbreviated
Z-Image Turbo
- + Highly accurate text rendering for the date, time, and location
- + Rich, cinematic lighting on the jack-o-lantern and trees
- + Better inclusion of requested elements like the thorn border
- − Minor typo in location ('Archves' instead of 'Arches')
- − The 'small scroll banner' is applied more like a page element than a functional banner for the specific quote
Verdict: Z-Image Turbo is the clear winner as it successfully rendered almost all the complex text requirements, including the date and time, whereas HiDream I1 Full generated nonsensical 'Yootmber' text. Z-Image Turbo also delivered a more atmospheric, cinematic look that better matched the 'vintage gothic' prompt description.
Bald man challenge
Image Editing“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
AI Judge Analysis
HiDream I1 Full
- + Attempts to add longer hair texture reaching behind the neck.
- − Fails to preserve the source image, completely changing the background and facial features.
- − Adds a bizarre, non-requested black object across the face.
- − Does not provide a 'full, thick head of hair' on the top of the scalp, showing a buzz-cut pattern instead.
Z-Image Turbo
- + Excellent source preservation, keeping the background, lighting, and jacket identical to the original.
- + Preserves the facial features and glasses accurately.
- − Fails to provide a 'full, thick head of hair', only adding a very slight stubble increase.
- − The hairline edit is extremely minimal and does not fulfill the core prompt instruction.
Verdict: HiDream I1 Full fails significantly as an image editor by completely changing the environment and adding an inexplicable black artifact to the subject's face. While Z-Image Turbo followed the 'source preservation' aspect of the prompt perfectly, it failed to actually provide the requested 'thick head of hair,' resulting in a very subtle change that mostly missed the goal. However, Z-Image Turbo is preferred for maintaining the integrity of the original photo.
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 Full
- + Excellent text rendering with clean, bold typography.
- + Includes a much more detailed and appealing variety of sushi types.
- + Very strong adherence to the isometric perspective and 3D diorama request.
- − Failed to include the requested flag icon.
- − The wasabi detail is slightly floating or poorly integrated into the tray texture.
Z-Image Turbo
- + Includes a flag icon as requested.
- + Clean, soft textures that match the cartoon miniature aesthetic well.
- − Incorrectly displayed the flag of China instead of Japan.
- − Only shows a single piece of sushi, making the scene feel empty and less like a diorama.
- − Slight text alignment issues with the flag overlapping the frame of the composition.
Verdict: HiDream I1 Full is the clear winner for its superior composition, variety of sushi, and professional text rendering, despite missing the small flag icon. Z-Image Turbo followed the flag prompt but provided the wrong country's flag (China) and produced a much simpler, less balanced image.
Over-the-top cartoon caricature
Editing“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”
AI Judge Analysis
HiDream I1 Full
- + Strong adherence to the 'caricature' art style request.
- + Includes clear visual elements for all requested themes including a dog and sports on the TV.
- + Maintains recognizable facial features from the source while exaggerating them.
- − Misses an obvious hockey stick or equipment, showing generic sports on a background screen instead.
- − The hand connection to the dog is slightly messy.
Z-Image Turbo
- + Excellent preservation of the source image's identity and realism.
- + Subtly includes a small dog in the background.
- − Completely fails to apply the requested 'caricature' or 'exaggerated' style.
- − Fails to incorporate the TV show anchor or hockey themes.
- − The edit is so minimal it is barely noticeable.
Verdict: HiDream I1 Full followed the complex prompt by transforming the image into a caricature and incorporating the requested themes of dogs and TV broadcasting. In contrast, Z-Image Turbo largely ignored the stylistic and thematic instructions, providing a nearly identical photo to the source with only a tiny dog added. HiDream I1 Full is the clear winner for actually performing the requested image editing task.
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 Full
- + Excellent fur detail and texture
- + Strong lighting effects with clear God rays and golden hour ambiance
- − Failed to include the baby bunny
- − Included two kittens instead of the requested single kitten
- − The style feels slightly more digital/artificial rather than hyper-photorealistic
Z-Image Turbo
- + Perfectly included all requested animals: puppy, kitten, bunny, and fox
- + Natural, life-like poses that suggest movement and playfulness
- + Superior aesthetic that captures the requested 'wholesome' vibe effectively
- − Minor anatomy issue with the puppy's paw touching the bunny
- − The 'God rays' are very subtle compared to the other model
Verdict: Z-Image Turbo is the clear winner as it successfully included all four requested animals, whereas HiDream I1 Full missed the rabbit entirely and duplicated the kitten. Z-Image Turbo also achieved a more natural, photorealistic look with dynamic poses that fit the 'tumbling' and 'chasing' part of the prompt much better than the static arrangement in the other image.
Studio Ghibli Anime Style
Editing“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”
AI Judge Analysis
HiDream I1 Full
- + Strong adherence to the Studio Ghibli art style with characteristic linework and facial features.
- + Captures the requested soft pastel colors and hand-painted aesthetic.
- + Correctly interprets the 'distracted boyfriend' meme composition while stylizing the characters.
- − Changes the emotional tone from conflict/confusion to a more generic happy anime scene.
- − Loss of some original features like the specific plaid pattern of the shirt for more generic floral patterns.
Z-Image Turbo
- + Perfectly preserves the source image composition and identity of the subjects.
- + Maintains the specific emotional expressions of the original meme.
- − Completely fails to apply the requested Studio Ghibli illustration style.
- − The output is virtually indistinguishable from the original photograph with only minor color adjustments.
Verdict: HiDream I1 Full successfully transformed the image into a Studio Ghibli-inspired illustration as requested, effectively changing the medium while keeping the core arrangement. In contrast, Z-Image Turbo failed to apply any significant stylistic change, keeping the image in a photographic format. HiDream I1 Full is the clear winner for following the artistic instruction despite some creative liberties with the facial expressions.
Golden Hour Stroll
Image Editing“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
AI Judge Analysis
HiDream I1 Full
- + Successfully captures a high-energy, lively feel as requested.
- + Effective use of many flying leaves to convey motion.
- + Hair shows significant motion.
- − Completely loses the golden retriever from the source image.
- − The person is significantly altered and no longer resembles the source.
- − Artificial-looking leaves and low-resolution artifacting throughout.
Z-Image Turbo
- + Excellent source preservation, keeping the original person, dog, and background almost perfectly intact.
- + Successfully adds subtle wind effects to the hair.
- + Maintains the composition of the original photo without hallucinations.
- − The 'flying leaves' are very sparse and small, making them easy to miss.
- − Less dynamic than requested, feeling more like a slight breeze than an energetic scene.
Verdict: HiDream I1 Full creates a more 'energetic' image by adding motion blur and many leaves, but it fails as an editing tool because it completely removes the dog and replaces the subject. Z-Image Turbo captures the prompt's intent while perfectly preserving the identity of the person and the presence of the dog, making it the superior choice for a source-based image edit.
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 Full
- + Successfully included the 'Est. 1720' banner as requested.
- + Clean vector style with a balanced color palette.
- + Creative incorporation of a coffee cup icon within the cloche.
- − Completely missed the primary text 'Caffè Florian'.
- − The steam effect is slightly clunky and less elegant than Model B.
Z-Image Turbo
- + Excellent typography that accurately renders 'Caffè Florian' with a professional vintage feel.
- + Stronger minimalist aesthetic that fits a real-world logo application.
- + Symmetry and spacing are well-balanced.
- − Failed to include the requested banner, opted for simple text lines instead.
- − The steam appears to be coming out of the handle rather than the food/drink.
Verdict: While HiDream I1 Full followed the banner requirement better, it failed the most critical part of a logo prompt: the name of the establishment. Z-Image Turbo produced a superior, more professional logo with perfectly rendered text, making it the more useful output despite missing the banner 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 Full
- + Excellent adherence to the navy-heavy color palette
- + Clean, high-quality vector aesthetics with consistent strokes
- + Captures the spirit of NASA mission patches in the header typography
- − Nonsensical text labels that seem to include prompt words like 'ICON' and 'UNG'
- − Fails to include all 6 requested steps, focusing on vertical launch imagery
- − Misinterprets the Saturn V rocket by adding unnecessary shuttle-like wings
Z-Image Turbo
- + Successfully includes the lunar module and lunar surface as requested
- + Better attempts at logical step sequencing and corresponding icons
- + Clear, legible typography for the body text
- − Major spelling errors in the main title ('APOLIO E 11') and subtitles ('Translurian')
- − Much simpler icon design compared to the requested modern vector style
- − Incorrect color for the lunar module (orange/yellow instead of muted red or gray)
Verdict: HiDream I1 Full produces a much more visually polished vector illustration that perfectly captures the requested NASA aesthetic, but it fails on instructional content and text accuracy. Z-Image Turbo does a better job of including the requested technical steps (like the lunar module), but is marred by a massive spelling error in the primary header and a more juvenile icons style. HiDream I1 Full is the likely winner for its superior visual quality and adherence to the design style prompt despite the poor text logic.
Explore each model
Tongyi-MAI's 6-billion parameter distilled text-to-image model optimized for speed, achieving high-quality generation in 8 steps or fewer with support for bilingual text rendering