Head to head
Esc

Models · slot A

to navigate to pick

FLUX.2 [dev] Turbo fal HiDream I1 Full HiDream AI

Settled by community votes across 16 shared challenges, with an AI judge weighing in on each.

FLUX.2 [dev] Turbo

27.9 arena score

#3 of 62 in Text-to-Image

Top 3 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

FLUX.2 [dev] Turbo

0.0%

win rate

Ties

0.0%

HiDream I1 Full

100.0%

win rate

0.0% 0.0% ties 100.0%
Shared challenges 16

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent realism with believable glass textures and reflections.
  • + Perfect spatial arrangement of the cube, sphere, and book.
  • + Natural integration of the green plant visible through the glass panes.
  • The sphere appears slightly dark, almost indigo rather than a bright blue.

HiDream I1 Full

  • + Strong adherence to the 'blue sphere' instruction with a vibrant color.
  • + Good lighting effect from the left side as requested.
  • + Clean, simple composition.
  • The blue sphere is floating unnaturally inside the cube.
  • The perspective of the cube's base is slightly warped compared to the table surface.
  • Reflections in the glass are a bit muddy and lack the clarity of Model A.

Verdict: FLUX.2 [dev] Turbo produces a significantly more realistic image with superior handling of glass physics and object placement. While HiDream I1 Full follows the prompt's color requirements well, it fails on basic physics by making the sphere float, and it lacks the fine detail and textural depth found in the FLUX output.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the 'motion blur from passing cars' prompt requirement.
  • + Highly realistic skin textures and mechanical details on the bicycle.
  • + Effective use of reflections and wet pavement to create a cinematic atmosphere.
  • The man appears to be kneeling directly on wet asphalt without a mat which feels slightly unnatural.
  • Slightly cluttered background compared to Model B.

HiDream I1 Full

  • + Strong bokeh effect and shallow depth of field.
  • + Good focus on the subject with a clean, centered composition.
  • + Realistic reflections on the very wet street surface.
  • Failed to include the requested motion blur for the cars.
  • The man's scale relative to the bicycle is slightly off, appearing too large for the frame.
  • The bicycle's anatomy is simplified and less realistic than in Model A.

Verdict: FLUX.2 [dev] Turbo significantly outperformed HiDream I1 Full by following every specific prompt detail, particularly the motion blur and the mechanical realism of the bicycle repair. HiDream I1 Full produced a high-quality image but missed the motion blur requirement and featured a less convincing interaction between the man and the bike.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Natural and lifelike skin textures and eyes.
  • + Exceptional detail on the engraved armor and leather straps.
  • + Superior integration of braids and beads with realistic physics.
  • The torch in the background is a bit blurry and simplified.
  • The facial hair is more of a light stubble than a full beard if that was expected, though fits the prompt.

HiDream I1 Full

  • + Strong bokeh effect with large, pleasant light circles.
  • + Dynamic facial expression and intense gaze.
  • Skin texture appears overly sharpened and plastic-like.
  • Braids look stiff and unnaturally symmetrical.
  • The armor engraving lacks the depth and realism of the first image.

Verdict: FLUX.2 [dev] Turbo produced a much more realistic and high-fidelity image, specifically excelling in the fine textures of the leather, metal, and human skin. HiDream I1 Full looks more like an artificial game render with over-sharpened details and less believable lighting. FLUX.2 better captured the 'battle-worn' aesthetic through subtle, realistic grime and complex armor patterns.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent typography with clean sans-serif fonts and readable pricing.
  • + Features a distinct grid layout for photography as requested.
  • + Includes vibrant color accents and professional graphic design elements.
  • Prices are unrealistically high ($160-$300 for a pizza).
  • One photo depicts a jar of pickles/vegetables which feels slightly out of place for the specific cuisine.

HiDream I1 Full

  • + Matches the minimalist request with a very simple, clean white background.
  • + Coherent food photography style across all images.
  • Poor typography with overlapping characters and inconsistent font weights.
  • Lacks the 'vibrant accents' requested in the prompt, appearing quite plain.
  • The 'Appetizers / Pizza' section is merged, whereas the prompt asked for sections for each.

Verdict: FLUX.2 [dev] Turbo produces a much more professional and usable design that follows the specific layout and accent requests of the prompt. While HiDream I1 Full captures the minimalist essence, its poor text rendering and failure to include vibrant accents make it less effective as a graphic design piece.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Perfectly renders every specific menu item and price exactly as requested in the prompt.
  • + The chalk texture is highly realistic with believable smudges and variations in stroke pressure.
  • + Excellent composition that feels like a natural photograph of a café interior.
  • A small repetition of the word 'Cookies' at the bottom of the list.

HiDream I1 Full

  • + Successfully renders the title and date with clear handwriting.
  • + Includes a charming chalk-style illustration of a coffee cup.
  • Completely fails to include the requested menu items, providing gibberish text instead.
  • The text appears more like a digital font overlay than natural chalk handwriting.
  • All prices are generic or nonsensical ($50, $2000) rather than the prompt's specifications.

Verdict: FLUX.2 [dev] Turbo followed the prompt almost perfectly, accurately spelling complex menu items like 'Truffle Mushroom Risotto' in a convincing chalk style. In contrast, HiDream I1 Full failed significantly on prompt adherence, replacing the requested items with illegible text and ignoring the specific prices and food descriptions.

The Reversed Rodeo

Text-to-Image

“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent realism and cinematic lighting with detailed textures on the spacesuit and horse fur.
  • + Complex composition featuring multiple planets and realistic debris in the foreground.
  • + High structural coherence in the horse's anatomy and the astronaut's gear.
  • Failed to follow the specific spatial instruction 'horse on top, not vice versa'.
  • The image is a standard interpretation of the common astronaut-on-horse trope.

HiDream I1 Full

  • + Strong dynamic sense of motion with the horse's pose in space.
  • + Clean, high-contrast aesthetic that feels surreal and dreamy.
  • Failed to follow the specific spatial instruction 'horse on top, not vice versa'.
  • Anatomical issues with the horse's legs appearing slightly distorted and poorly attached.
  • Lower level of fine detail compared to the competing model.

Verdict: Both models completely ignored the difficult negative/spatial constraint to place the horse on top of the astronaut, instead defaulting to the common 'astronaut riding a horse' trope. FLUX.2 [dev] Turbo is the clear winner due to its significantly higher visual quality, realistic lighting, and intricate details in the space environment, whereas HiDream I1 Full shows some anatomical distortion and less sophisticated textures.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent photorealism in lighting and textures
  • + Accurately places the passenger in the back seat as requested
  • + Detailed capybara fur and realistic 'hands' on the wheel
  • The passenger is sitting in the middle of the front seat/back seat gap rather than clearly behind the driver
  • The steering wheel looks a bit thin and digital

HiDream I1 Full

  • + Natural composition and clear depth of field
  • + Clean text rendering on the hat
  • + Good adherence to the calm expression prompt
  • The human passenger is sitting in the front passenger seat, failing the 'back seat' part of the prompt
  • The capybara's hands/paws are slightly distorted on the steering wheel
  • Lacks the gritty, realistic atmospheric lighting of the alternative

Verdict: FLUX.2 [dev] Turbo succeeded in placing the passenger in the back seat and providing a much more convincing photorealistic aesthetic, including the exterior NYC lighting. HiDream I1 Full failed to place the businesswoman in the back seat, positioning her in the front instead, and had less realistic integration of the animal and its clothing.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full
0% wins 0% ties 100% wins

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Perfect text rendering of all requested details including specific dates and addresses.
  • + Excellent atmospheric lighting and cohesive vintage gothic aesthetic.
  • + Frame composition with thorns and webs matches the prompt accurately.
  • The jack-o-lantern is slightly off-center.

HiDream I1 Full

  • + Includes all thematic elements like bats, trees, and jack-o-lantern.
  • + The scroll banner design is visually distinct and prominent.
  • Significant text errors and hallucinations in the date, time, and location segments.
  • Missed the specific requested title 'Halloween Party Invitation', shortening it to 'Halloween Party'.
  • Lower visual fidelity with less realistic textures on the paper and trees.

Verdict: FLUX.2 [dev] Turbo followed the complex text instructions perfectly, including the specific date and location which are essential for an invitation. HiDream I1 Full failed on most of the specific text requirements, resulting in gibberish for the event details. FLUX.2 also provided a much more polished and cinematic atmosphere that better fits the 'vintage gothic' request.

Bald man challenge

Image Editing
Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After
FLUX.2 [dev] Turbo
Before After
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent preservation of the source image's facial features and background
  • + Highly realistic hair texture that blends naturally with the existing beard
  • + Perfectly followed the instruction to provide a full head of hair while maintaining the original lighting
  • The hairline transition on the forehead is slightly soft compared to the rest of the image

HiDream I1 Full

  • + Added some hair texture to the back of the head
  • Failed to provide a 'full head of hair', instead giving a thin ponytail and keeping the crown bald
  • Completely changed the person's face and identity
  • Introduced bizarre hallucinations like a black plastic object coming out of the mouth and a completely different background

Verdict: FLUX.2 [dev] Turbo successfully performed the edit by adding a realistic, full head of hair while keeping the subject's face and environment identical to the source. HiDream I1 Full failed on nearly every metric, altering the person's face, changing the background, and adding strange artifacts while neglecting the primary request for a full head of hair.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent text rendering and placement according to instructions.
  • + Includes the requested small flag icon correctly.
  • + Highly realistic textures on the salmon and rice grains.
  • The camera angle is slightly lower than the requested 45-degree top-down isometric view.

HiDream I1 Full

  • + Perfect 45-degree isometric perspective and dioramic base.
  • + Clean cartoon/3D aesthetic consistent with the prompt style.
  • + Good variety of sushi types on the platter.
  • Missed the request for the small flag icon.
  • The 'SUSHI' text is very small and slightly off-center compared to 'JAPAN'.
  • Low-resolution textures on the background with some visible noise/artifacts.

Verdict: FLUX.2 [dev] Turbo followed the complex text and icon instructions perfectly, producing a highly clear image with impressive material realism. While HiDream I1 Full captured the better isometric composition and diorama base, it failed to include the flag icon and had less refined text rendering. FLUX.2 is preferred for its superior adherence to the specific typography and lighting details.

Over-the-top cartoon caricature

Editing
Edit instruction

“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.”

Source
FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to all prompts: TV anchor desk, hockey rink background, and multiple dogs.
  • + Great caricature style with stylized, exaggerated proportions.
  • + Maintains the likeness of the source subject while translating it to a cartoonish medium.
  • The text on the desk is nonsensical gibberish.
  • The composition is a bit crowded with many competing elements.

HiDream I1 Full

  • + Successfully creates a high-quality caricature face with a clear likeness to the original photo.
  • + Incorporates a dog into the main composition naturally.
  • Completely fails the 'hockey' requirement of the prompt.
  • The 'TV anchor' profession is very subtly implied by a small TV in the background rather than being the central theme.
  • The dog's placement on the shoulder is a bit awkward and lacks distinct context.

Verdict: FLUX.2 [dev] Turbo followed the complex instructions much better, creating a cohesive scene that merged the TV anchor profession, hockey, and dogs into a single, humorous caricature. HiDream I1 Full produced a good caricature likeness but ignored the 'hockey' instruction and placed the subject in a generic living room context rather than at a TV anchor desk.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent adherence to the full animal list including the bunny.
  • + Superior anatomical realism and fur texture.
  • + Expertly captures the 'god rays' and 'dew sparkles' requested in the prompt.
  • The kitten's facial expression is slightly frantic.

HiDream I1 Full

  • + Warm and pleasant lighting effect.
  • + Cute, stylized interpretation of the animals.
  • Failed to include the baby bunny as requested.
  • Replaced the bunny with a second kitten, missing the prompt requirements.
  • Noticeable anatomy issues with the animals' legs and paws blending into the grass.

Verdict: FLUX.2 [dev] Turbo followed the prompt much more accurately by including all four distinct animal types, whereas HiDream I1 Full missed the rabbit entirely and included two kittens instead. Furthermore, FLUX.2 exhibited far superior realism and finer details in the fur, background elements like dew, and overall spatial coherence.

Studio Ghibli Anime Style

Editing
Edit instruction

“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”

Source
FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent preservation of the original poses and facial expressions.
  • + Captures the wash-style, hand-painted aesthetic of Studio Ghibli backgrounds perfectly.
  • + Maintains the identifiable features and clothing of the people in the meme.
  • The color palette is a bit more desaturated than classic Ghibli vibrant scenes.

HiDream I1 Full

  • + Successfully adopts a cute anime art style.
  • + Adds fun Ghibli-esque details like the floral patterns and straw hat.
  • Fails to preserve the core 'distracted boyfriend' dynamic as the man is smiling at the woman in red rather than being caught looking.
  • Significant changes to clothing and character features make the source image unrecognizable.

Verdict: FLUX.2 [dev] Turbo is the clear winner as it successfully translates the meme into a new style while maintaining the exact composition and emotional intent of the original photo. HiDream I1 Full changes the narrative of the image, showing the man smiling happily at the woman in front rather than looking back in secret, which misses the point of the edit.

Golden Hour Stroll

Image Editing
Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

Before After
FLUX.2 [dev] Turbo
Before After
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Excellent preservation of the source image's identity, dog, and background.
  • + Directly follows instructions by adding wind-blown hair and flying leaves.
  • + Maintains high resolution and clarity without distorting features.
  • The wind effect on the hair is a bit symmetrical and looks slightly artificial.
  • Some leaves appear to be floating on top of the person rather than in 3D space.

HiDream I1 Full

  • + Features a very high level of dynamic energy and movement.
  • + Naturalistic rendering of wind-blown hair and falling leaves.
  • Fails as an image editor by completely removing the dog from the scene.
  • Drastically changes the identity of the person and the background elements.
  • Does not preserve the source image's composition.

Verdict: FLUX.2 [dev] Turbo successfully performed the edit while keeping the original image intact, adding the requested wind and leaf effects accurately. HiDream I1 Full failed the underlying task of image editing, as it generated an entirely new image that removed the dog and changed the person, despite capturing a higher level of 'energy'.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Perfect text rendering for both the main name and the banner date.
  • + Excellent use of texture and vintage paper effects that match the prompt.
  • + Superior vector-style illustration with clean lines and classic shading.
  • The 'Est. 1720' text is slightly off-center within the banner.

HiDream I1 Full

  • + Creative integration of a coffee cup icon inside the cloche.
  • + Clean, high-contrast minimalist flat design.
  • + Adheres well to the warm brown and cream color scheme.
  • Failed to include the primary text 'Caffè Florian'.
  • Missing the 'subtle texture' requested in the prompt.
  • The steam lines and cloche shape are slightly unbalanced

Verdict: FLUX.2 [dev] Turbo followed all instructions, including the specific name 'Caffè Florian', which HiDream I1 Full omitted entirely. FLUX.2 also provided a much more convincing vintage texture and professional vector aesthetic, whereas HiDream I1 Full produced a generic icon with much less detail and brand identity.

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.”

FLUX.2 [dev] Turbo
HiDream I1 Full

AI Judge Analysis

FLUX.2 [dev] Turbo

  • + Includes all six requested steps of the mission
  • + Excellent text rendering for names and stages
  • + Highly accurate iconography representing the Saturn V and Lunar Module
  • Layout is a bit cluttered with redundant 'Translunar' and 'Landing' labels
  • Literal interpretation of the prompt included the word 'icon' in the 'Saturn V icon' label

HiDream I1 Full

  • + Clean, minimalist flat-vector aesthetic
  • + Good adherence to the requested NASA-inspired color palette
  • + High-quality vector shapes and lines
  • Fails to include all six requested steps, stopping after the initial stages
  • Nonsensical text additions like '+ UNG' and 'UNLY'
  • Includes a generic ringed planet instead of the Moon

Verdict: FLUX.2 [dev] Turbo followed the complex instructions much more effectively, including every specific step and icon requested, while maintaining very high legibility. HiDream I1 Full captured the requested art style well but failed significantly on the logic of the infographic, missing several mission steps and rendering inaccurate celestial bodies.

Next steps

Explore each model