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
Seedream 4.5
#9 of 62 in Text-to-Image
Where the votes landed
HiDream I1 Full
0%
win rate
Ties
0%
Seedream 4.5
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 soft window lighting with realistic diffusion on the table
- + Good spatial arrangement of elements and depth of field
- + Realistic reflections of the sphere on the inner glass surfaces
- − The blue sphere is levitating unnaturally in the center of the cube
- − The back edge of the table is cut off by the framing
Seedream 4.5
- + Natural physics with the sphere resting on the bottom of the cube
- + Clearer visibility of the plant through the glass panels
- + Better framing of the tabletop and the subject
- − The red book has a slight doubling/artifact on the right-hand spine edge
- − The lighting feels slightly harsher compared to the 'soft' request
Verdict: Both models followed the complex spatial instructions perfectly. Seedream 4.5 is the preferred winner because it placed the sphere on the floor of the cube realistically, whereas HiDream I1 Full showed the sphere floating, and Seedream 4.5 also provided a more complete composition of the table.
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 unavailable for this challenge.
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
- + Excellent adherence to the 'beads' and 'braids' prompt with intricate hair styling
- + Highly detailed engraving on the plate armor
- + Intense, lifelike gaze and clear facial expression
- − The scars look a bit like red ink or surface paint rather than deep, aged skin tissue
- − Lighting feels slightly flat compared to the dramatic expectations of 'torchlight'
Seedream 4.5
- + Exceptional photographic realism with natural skin texture, dirt, and realistic scarring
- + Superior cinematic lighting that realistically reflects warm torchlight off the facial planes and gold armor
- + Perfect execution of shallow depth of field and bokeh sparks
- − The braids are less prominent than requested in the prompt
- − The beads in the hair look a bit uniform and static
Verdict: While HiDream I1 Full creates a striking and detailed symmetry, Seedream 4.5 is the clear winner due to its superior realism and professional cinematic lighting. Seedream 4.5 captures the 'faint scars and dirt' with much more nuance and provides a more believable photographic texture on the leather and metal.
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
- + Successfully renders very bold and legible header text.
- + Food photography of the pizza is high quality and appetizing.
- − Redundant imagery, showing four nearly identical pizzas across different sections.
- − Layout is a bit clunky with text overlapping or crowding the image boundaries.
Seedream 4.5
- + Excellent logical layout with distinct sections for Appetizers, Pizza, and Mains.
- + Effective use of vibrant color accents through the image borders.
- + Diverse food photography that matches the specified categories.
- − Text rendering under sections is gibberish and repetitive.
- − Prices listed are unrealistic for a casual dining menu ($99 for a main).
Verdict: Seedream 4.5 is the preferred model because it better interpreted the 'grid' and 'sections' request by providing diverse food images that actually correspond to 'Appetizers', 'Pizza', and 'Mains'. While HiDream I1 Full has cleaner text rendering in the headers, its repetitive use of the same pizza image for every section fails the professional layout criteria.
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
- + Successfully renders the date April 30, 2026
- + Includes a clear chalkboard background and frame
- − Fails to include the specific requested menu items
- − Text is mostly gibberish with inconsistent pricing
- − Handwriting looks like a digital font rather than authentic smudge-prone chalk
Seedream 4.5
- + Perfectly adheres to the requested menu items and descriptions
- + Captures an incredibly realistic chalk texture with dust and smudges
- + Maintains high text legibility while looking genuinely handwritten
- − Repeats the price '$24' unnecessarily on two lines for the first item
Verdict: Seedream 4.5 captures the prompt's requirements with high accuracy, rendering the specific menu items perfectly with a very realistic chalk aesthetic. HiDream I1 Full fails to follow the content instructions, producing random gibberish text and a more artificial font-like appearance.
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
- + Excellent anatomical rendering of the horse and astronaut
- + Dynamic sense of motion with the horse in a gallop
- + High clarity and cinematic lighting against the Earth backdrop
- − Followed the standard 'astronaut on horse' trope despite the specific negative constraint
Seedream 4.5
- + Vibrant color palette with nebula effects
- + Good rendering of the gold visor texture
- + Cinematic composition with a vertical 'rearing' pose
- − Failed to follow the specific instruction for the horse to be on top
- − Anatomy issues with the astronaut's legs and the way they merge into the horse
Verdict: Both HiDream I1 Full and Seedream 4.5 failed the negative constraint to have the 'horse on top', instead producing traditional 'astronaut riding horse' images. HiDream I1 Full is the superior image due to its cleaner anatomical details, more realistic horse pose, and overall professional cinematic rendering compared to the muddy leg details in Seedream 4.5.
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 fur texture and lighting on the capybara.
- + Sharp focus on the passenger's expression and phone detail.
- + Vibrant bokeh effect in the background.
- − The capybara has five-fingered ape-like hands instead of paws.
- − The seat configuration looks slightly cramped and anatomically confusing near the door.
Seedream 4.5
- + Better perspective that shows more of the taxi's front window and dashboard.
- + More accurate depiction of capybara paws on the steering wheel.
- + Captures a very authentic 'bored' expression on the passenger.
- − The passenger's hands and fingers holding the phone are poorly rendered and distorted.
- − Slightly less 'photorealistic' skin texture on the human compared to Model A.
Verdict: HiDream I1 Full produces a more polished and high-end photographic look with better clarity, but Seedream 4.5 captures the requested perspective and animal features more accurately. Seedream 4.5 is the preferred choice because the capybara's anatomy is more believable, even though the human passenger's hands suffer from typical AI artifacts.
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
- + Successfully included the spooky border with thorns and webs in the layout
- + The jack-o-lantern has a high-quality glow and detail
- + The central scroll banner contains the correct text
- − Failed significantly with the specific event details, generating nonsensical text instead of the requested date and time
- − Missing the full 'Invitation' word in the main title
- − The illustration style is more clip-art than cinematic
Seedream 4.5
- + Excellent text rendering covering all requested event details including the date, time, and location
- + Strong cinematic lighting and atmospheric moody sky as requested
- + High adherence to the central scroll banner and invitation text
- − The thorns at the bottom look more like modern barbed wire than vintage gothic thorns
- − The dark parchment element is not visually distinct, appearing more like a standard photo overlay
Verdict: Seedream 4.5 is the clear winner because it correctly follows the complex text instructions for the event details, whereas HiDream I1 Full produces nonsensical gibberish for the date and location. Seedream 4.5 also captures the cinematic lighting and polished atmosphere much better than the flatter, more graphic design style of HiDream I1 Full.
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 volume to the back of the head
- − Fails to add hair to the top of the head, leaving the subject bald
- − Introduces nonsensical artifacts like the black object floating in the beard
- − Completely changes the background, facial features, and adds a backpack, failing source preservation
Seedream 4.5
- + Perfectly adheres to the prompt by adding a realistic, thick head of hair
- + Flawlessly preserves all facial features, glasses, clothing, and background from the source image
- + Maintains the lighting and texture consistency of the original photograph
- − None observed
Verdict: Seedream 4.5 is the clear winner as it perfectly executed the requested edit while maintaining 100% fidelity to the unedited parts of the source image. In contrast, HiDream I1 Full failed to add hair to the top of the head, hallucinated strange objects into the beard, and completely altered the subject's face and environment.
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 typography style that blends with the 3D aesthetic.
- + High-quality textures on the sushi and rice grains.
- + Strong perspective and layout on the diorama base.
- − Missed the request for a flag icon.
- − The text is slightly off-center globally.
Seedream 4.5
- + Followed all prompt instructions including the small flag icon.
- + Achieved a very realistic PBR material look on the salmon and plate.
- + Clean, minimalist composition that feels miniature.
- − The text 'JAPAN' is not perfectly centered and looks like a generic 2D overlay.
- − The texture of the diorama base is a bit coarse compared to the 'soft refined' request.
Verdict: Seedream 4.5 followed the prompt more literally by including the flag icon and delivering a high-quality PBR render, though its typography integration is weaker. HiDream I1 Full created a more cohesive artistic piece with better-integrated text, but missed the flag requirement. Seedream 4.5 is the preferred choice for strictly following the complex list of requirements.
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
- + Successfully captures the subject's likeness in a classic caricature drawing style.
- + Includes subtle hockey references on the background TV screen.
- − Misses the 'hockey' requirement almost entirely, failing to make it a prominent feature.
- − The caricature style feels a bit aged and less polished than the competitor.
Seedream 4.5
- + Excellent adherence to all prompt elements: includes a microphone, hockey stick, gloves, puck, and a dog.
- + Strong caricature proportions with a large head while maintaining the subject's facial features professionally.
- + Cleverly incorporates the anchor desk and news studio setting.
- − The transition from the 3D-style character to the photorealistic dog and setting is slightly jarring.
- − The hands on the desk are slightly small and less detailed.
Verdict: Seedream 4.5 is the clear winner as it fulfilled every part of the prompt, including the hockey and news anchor elements which HiDream I1 Full largely ignored. Seedream 4.5 also maintained a better visual balance between the original source image's features and the caricature-style exaggeration.
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 texture and sharpness on the main subjects.
- + Features more butterflies scattered throughout the scene.
- + Vibrant color palette with clean, defined lighting.
- − Completely missed the 'baby bunny' requirement, providing a second kitten instead.
- − The characters are posing statically rather than 'tumbling and chasing'.
- − The eyes look somewhat artificial and overly glassy.
Seedream 4.5
- + Successfully included all four requested animals: puppy, kitten, bunny, and fox.
- + Better adherence to the dynamic action of 'tumbling' and 'chasing' with motion-blur paws.
- + Excellent implementation of 'god rays' and 'dew sparkles' (bokeh drops).
- − The fox's eyes appear slightly distorted and mismatched in size.
- − The kitten's tail looks slightly elongated and anatomically awkward.
- − The lighting on the dog's face is a bit flat compared to the background intensity.
Verdict: Seedream 4.5 is the clear winner because it correctly followed the prompt's instruction to include a bunny, whereas HiDream I1 Full omitted the bunny in favor of a second kitten. Seedream 4.5 also captured the dynamic energy of the animals playing and tumbling, while HiDream I1 Full created a static, posed portrait.
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
- + Successfully captures a vibrant anime aesthetic with detailed character designs.
- + Updates the setting with interesting Ghibli-esque signage and floral patterns.
- − Fails to preserve the core narrative and composition of the source meme.
- − Changes the man's expression from a distractible look to a pleasant smile, losing the point of the original image.
Seedream 4.5
- + Excellent source preservation, maintaining the exact poses and composition of the original meme.
- + Accurately applies the Studio Ghibli 'watercolor' background style and character linework.
- + Preserves the specific facial expressions (pout, distraction, anger) while translating them to anime.
- − The transition between the sharp characters and the soft watercolor background is a bit harsh in some areas.
Verdict: Seedream 4.5 is the clear winner as it successfully performs the edit while maintaining the structural integrity and meaning of the source meme. HiDream I1 Full reimagines the scene entirely, losing the iconic 'distracted boyfriend' composition, whereas Seedream 4.5 perfectly translates the original photo's expressions and layout into a convincing hand-painted Ghibli style.
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 highly energetic and lively feeling through the subject's pose.
- + Excellent rendering of flying leaves with a convincing sense of depth.
- − Fails to preserve the original image, removing the dog and changing the subject's face entirely.
- − The landscape elements like the bridge and bushes are significantly altered.
Seedream 4.5
- + Perfectly preserves the identity of the woman and the presence of the dog from the source image.
- + Accurately applies all requested edits including blowing hair and flying leaves while maintaining the original background.
- + Maintains the original composition and lighting.
- − The leash handle has been slightly malformed compared to the original.
- − The motion of the hair feels somewhat stiff on the left side.
Verdict: Seedream 4.5 is the clear winner because it functions as an actual image editing tool, successfully adding the requested wind and leaves while keeping the woman, the dog, and the environment identical to the source. HiDream I1 Full failed the fundamental task of image editing by generating an entirely new image that removed the dog and changed the woman's appearance, despite having high artistic quality on its own.
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
- + Clean vector-style execution with high contrast.
- + Accurate rendering of the banner and 'Est. 1720' text.
- − Completely failed to include the primary text 'Caffè Florian' from the prompt.
- − The cloche looks more like a bell or jar because of the weird cup cutout inside it.
Seedream 4.5
- + Successfully included all prompt text including 'Caffè Florian'.
- + Better representation of 'subtle texture' on the background.
- + Accurate retro cloche design and elegant typography.
- − The steam effect is slightly off-center compared to the cloche handle.
Verdict: Seedream 4.5 is the clear winner as it fulfilled the entire prompt, including the brand name which HiDream I1 Full omitted entirely. Seedream 4.5 also maintained a superior vintage aesthetic with better shading on the cloche and a more nuanced parchment-like texture.
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 primary title typography.
- + Attractive use of large-scale vector graphics.
- + Strong adherence to the requested color palette.
- − Failed to include 5 out of the 6 requested steps.
- − Text below icons includes prompt leakage like '+ ICON' and gibberish.
- − The layout is muddled and doesn't function as a sequential infographic.
Seedream 4.5
- + Followed the step-by-step instructions perfectly with all 6 stages represented.
- + Excellent text rendering for both labels and the names of the crew.
- + Clean, logical infographic layout that successfully conveys information.
- − The 'Descent' icon looks more like a general satellite than a lunar module.
- − Icon for 'Earth Orbit' shows a ring around the planet similar to Saturn.
Verdict: Seedream 4.5 is the clear winner for its superior prompt adherence, successfully mapping out all six requested stages of the Apollo 11 mission in a logical, readable format. HiDream I1 Full failed to follow the sequential instructions, only showing a few items and including literal prompt text like '+ ICON' in the final image.
Explore each model
ByteDance's latest image generation model unifying text-to-image and image editing in a single architecture, with improved text rendering and 30-40% faster generation than v4.0