Fast distilled version of Black Forest Labs' FLUX.2 [dev] optimized for speed and cost efficiency.
Settled by community votes across 16 shared challenges, with an AI judge weighing in on each.
FLUX.2 [dev] Flash
#5 of 62 in Text-to-Image
HiDream I1 Full
#60 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [dev] Flash
0%
win rate
Ties
0%
HiDream I1 Full
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
FLUX.2 [dev] Flash
- + Excellent realism in textures, particularly the wood grain and glass thickness.
- + Perfect adherence to the prompt with a natural-looking arrangement.
- + Sophisticated lighting and reflections within the glass cube.
- − The sphere is slightly dark, making it look almost navy instead of a bright blue.
HiDream I1 Full
- + Successfully captures the prompt elements in a clean composition.
- + Good usage of soft light and shadows on the wooden surface.
- − The blue sphere appears to be floating unnaturally in the center of the cube.
- − The perspective of the table and reflections are less convincing than the competitor.
- − Overall image quality is slightly softer and less detailed.
Verdict: FLUX.2 [dev] Flash produces a highly realistic image with superior textures and logical physics, where the sphere rests naturally on the bottom of the cube. HiDream I1 Full follows the prompt well but loses points for the levitating sphere and less sharp detail. FLUX.2 is the clear winner for its photographic quality and better spatial reasoning.
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
FLUX.2 [dev] Flash
- + Excellent adherence to technical prompts like motion blur and 50mm lens look
- + Highly realistic skin textures and believable hand interaction with tools
- + Stronger narrative composition with tools scattered on the pavement
- − The white car in the background is slightly distorted in its perspective
HiDream I1 Full
- + Beautiful wet pavement reflections
- + Atmospheric background with legible Japanese signage depth
- − Anatomical and structural issues: the man's legs pass through the bike frame
- − Lack of visible motion blur mentioned in the prompt
- − The bike design is physically nonsensical with missing components
Verdict: FLUX.2 [dev] Flash produces a much more coherent and realistic image, capturing the 'candid' and 'motion blur' requirements perfectly while maintaining physical integrity. HiDream I1 Full fails on anatomical logic, with the man's legs clipping through the bicycle frame and a lack of the requested motion effects.
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
FLUX.2 [dev] Flash
- + Natural and realistic skin textures including subtle pores and fine hairs
- + Highly intricate engraving patterns on the plate armor with realistic lighting
- + Excellent adherence to the 'beads in hair' request with various shapes and materials
- − The scars appear more like fresh bloody marks rather than the 'faint scars' requested
HiDream I1 Full
- + Strong cinematic lighting and high contrast between the subject and the background
- + Solid plate armor design that fits the paladin aesthetic well
- + Correct inclusion of braided hair and sparks
- − Skin texture appears overly sharpened and somewhat 'plastic' or AI-generated
- − The eyes have a slightly unnatural glow that lacks the requested lifelike depth
- − The armor engravings look a bit muddy and less detailed than Image A
Verdict: FLUX.2 [dev] Flash produces a much more realistic and detailed portrait with superior texture work on the leather, skin, and metal engraving. While HiDream I1 Full captures the cinematic atmosphere and the prompt elements well, it suffers from over-sharpening artifacts and less believable skin rendering compared to the high-fidelity output of FLUX.2.
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
FLUX.2 [dev] Flash
- + Excellent use of vibrant secondary colors to create a modern visual identity.
- + Successfully includes all requested layout features including sections for pizza, mains, and appetizers.
- + High-quality, distinct food photography that fits the grid layout perfectly.
- − Text rendering is messy with several spelling errors and illegible characters.
- − All the 'Appetizers' photos appear to be pizzas, showing a lack of semantic variety in the images.
HiDream I1 Full
- + Layout is very clean and adheres strictly to a minimalist professional aesthetic.
- + Text is much more legible and utilizes bold sans-serif fonts effectively.
- + Stronger source preservation of the white background for a print-ready look.
- − The grid of photos is less comprehensive compared to the prompt's request for sections.
- − Poor logic in pricing (using '200' and '500') and missing currency symbols.
- − All photos provided are identical in subject matter (pizza), ignoring the 'colorful food photos' variety requested.
Verdict: FLUX.2 [dev] Flash captures the creative spirit of the prompt much better, using vibrant accents and a complex grid that truly feels like a 'casual dining' menu. While HiDream I1 Full produces cleaner text and a simpler minimalist layout, it fails to provide the variety and vibrancy described in the prompt. FLUX.2 [dev] Flash is the winner for its better composition and closer adherence to the requested visual style.
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
FLUX.2 [dev] Flash
- + Excellent text rendering with no spelling errors in the main menu items.
- + Authentic chalk texture and smudging that mimics a real chalkboard.
- + Perfect adherence to the specific menu items requested in the prompt.
- − The 'Today's Specials' title is less elegant/cursive compared to the stylized font in the other model.
- − The background cafe details are a bit soft in focus.
HiDream I1 Full
- + Beautiful cursive calligraphy for the header.
- + High contrast and sharp lines make the board very legible.
- − Complete failure to follow the specific menu items requested, resulting in gibberish text.
- − The chalk effect looks more like a digital font overlay than hand-drawn chalk.
- − Incorrect formatting with two columns and a random coffee illustration not requested.
Verdict: FLUX.2 [dev] Flash followed the complex text instructions perfectly, including the specific food items and prices, while maintaining a very realistic chalk aesthetic. HiDream I1 Full failed significantly on prompt adherence, producing nonsensical text and ignored the specific menu items requested. FLUX.2 [dev] Flash is the clear winner for its superior text accuracy and realistic material texture.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
FLUX.2 [dev] Flash
- + Excellent high-detail rendering of the spacesuit and horse fur.
- + Beautiful cinematic lighting with realistic planetary depth.
- + Complex textures in the harness and background nebulae.
- − Failed the specific structural constraint of having the horse on top of the astronaut.
HiDream I1 Full
- + Dynamic composition with a strong sense of movement.
- + Clean, high-contrast visual style.
- + Good anatomy on the horse's legs during a gallop.
- − Failed the specific structural constraint of having the horse on top of the astronaut.
- − Less detailed textures on the spacesuit compared to Model A.
Verdict: Both FLUX.2 [dev] Flash and HiDream I1 Full failed to follow the specific surreal prompt instruction to place the horse on top of the astronaut, instead providing standard interpretations of an astronaut riding a horse. FLUX.2 [dev] Flash is the slightly better image overall due to its superior lighting, intricate textures, and cinematic atmospheric depth.
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
FLUX.2 [dev] Flash
- + Excellent photorealism with cinematic lighting and realistic textures.
- + Perfect adherence to prompt details including two front paws on the wheel and the specific bored expression of the woman.
- + Authentic New York taxi interior and exterior details like the top light and hood.
- − The capybara's claws are slightly exaggerated and scary looking.
HiDream I1 Full
- + Good composition with a clear side profile of both characters.
- + Text on the cap is legible and accurate.
- − The capybara's 'hands' appear morphed and unnatural on the steering wheel.
- − The woman appears to be sitting in the front passenger seat rather than the back seat as requested.
- − Lower overall photorealism compared to the competitor, with a slightly AI-generated sheen.
Verdict: FLUX.2 [dev] Flash is the clear winner as it followed every instruction, including the specific positioning of characters and paws. While HiDream I1 Full produced a decent image, it failed to place the woman in the back seat and struggled with the physical anatomy of the capybara's limbs.
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
FLUX.2 [dev] Flash
- + Excellent text rendering with almost perfect adherence to the specific date and location.
- + High visual quality with a cohesive moody gothic aesthetic.
- + Includes all requested elements like the thorn/web border and twisted trees.
- − One line of text includes gibberish symbols ('Hurk: 0369C') that was not in the prompt.
HiDream I1 Full
- + Clear, readable banner text and a well-designed central jack-o-lantern.
- + Strong composition with a clear 'parchment' effect.
- − Failed significantly on the event details, producing gibberish like 'Yootmber' and '30 cm'.
- − The art style is a bit more clip-art oriented and less 'cinematic' than requested.
- − Missing the specific location text requested.
Verdict: FLUX.2 [dev] Flash is the clear winner because it successfully followed the complex text instructions, including the specific date and location in NYC. While HiDream I1 Full captured the general layout, its failure to render the requested event details accurately makes it unusable as a functional invitation.
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
FLUX.2 [dev] Flash
- + Excellent preservation of the original person's facial features and identity.
- + Accurate adherence to the request for full, thick hair.
- + Seamless integration of the new hair with the existing lighting and background.
- − The hairline is slightly too uniform and straight across the forehead.
HiDream I1 Full
- + Added a new background with decent detail.
- − Failed to provide a full head of hair, keeping the subject mostly bald.
- − Completely changed the person's face and identity.
- − Added bizarre artifacts including a strange black object across the subject's face.
Verdict: FLUX.2 [dev] Flash successfully performed the edit while keeping the original person's face and background perfectly intact. HiDream I1 Full failed on nearly every metric, failing to add the requested hair, changing the subject's identity, and introducing nonsensical visual artifacts.
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
FLUX.2 [dev] Flash
- + Excellent typography rendering with the flag icon correctly placed
- + Sophisticated PBR textures that balance realism with a miniature aesthetic
- + Perfectly logical shadows and wood grain details on the diorama base
- − The sushi roll construction is slightly nonsensical with fish draped over a maki-style roll
- − Less variety in the types of sushi displayed compared to the other model
HiDream I1 Full
- + Strong 3D cartoon style with vibrant colors and diverse sushi types
- + Clean isometric representation with a well-defined diorama tray
- + Good clarity and soft lighting across the entire scene
- − Failed to include the required flag icon next to the text
- − The typography is slightly less refined with tighter spacing and less weight than requested
- − Plastic-like textures lack the realism of PBR materials specified
Verdict: FLUX.2 [dev] Flash followed the instructions more accurately, including both the text and the specific flag icon, while achieving a much more sophisticated 'realistic PBR' look. HiDream I1 Full produced a charming 3D cartoon scene with better sushi variety, but it missed the flag and the textures feel more like simple plastic than refined digital materials.
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
FLUX.2 [dev] Flash
- + Excellent adherence to all prompts including the TV anchor desk, the hockey rink setting, and multiple dogs.
- + Captures the subject's likeness very effectively in a classic bobblehead-style caricature.
- + High level of detail and humor, such as dogs wearing hockey gear and holding sticks.
- − The hands on the anchor are slightly small even for caricature proportions.
HiDream I1 Full
- + Successfully translates the subject's face into a stylized caricature illustration.
- + Maintains the outfit (denim shirt and black top) from the source image.
- − Fails to include any reference to hockey.
- − Weak interpretation of the 'tv show anchor' profession, relying only on a small screen in the background.
- − Artifact in the dog's mouth where a black object is clipping through.
Verdict: FLUX.2 [dev] Flash is the clear winner as it masterfully combined all three requested elements—TV anchoring, dogs, and hockey—into a cohesive and humorous rink-side news desk scene. HiDream I1 Full captured the facial likeness well but completely ignored the hockey requirement and provided a generic living room background that didn't read as a professional news set.
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
FLUX.2 [dev] Flash
- + Excellent execution of lighting with distinct god rays and dew sparkles mentioned in the prompt.
- + Superior anatomical details and fur textures that feel more photorealistic.
- + Included all requested animals, even if duplicated.
- − Duplicated the foxes and bunnies instead of focusing on a single group.
- − The kitten is quite small and tucked in the center, losing focus.
HiDream I1 Full
- + Clearer focus on individual animal faces.
- + Warm, vibrant color palette that captures the joyful vibe.
- − Failed to include the 'baby bunny' requested in the prompt.
- − The 'fox' has a slightly distorted, almost AI-painterly face compared to the realism of Model A.
- − Lighting is a bit overblown and lacks the specific 'god rays' and 'dew' details of Model A.
Verdict: FLUX.2 [dev] Flash followed the technical details of the prompt much more closely, successfully rendering the dew sparkles and god rays while including all four animal types. HiDream I1 Full missing the rabbit is a significant prompt adherence failure, and its overall quality leans more toward an illustration than the requested 'hyper-photorealistic' masterpiece. FLUX.2 is preferred for its superior texture and adherence to environmental lighting cues.
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
FLUX.2 [dev] Flash
- + Excellent preservation of the original subjects' poses and expressions
- + Accurate representation of the 'distracted boyfriend' meme composition
- + Beautiful hand-painted texture and soft pastel color palette
- − The background loses the street context in favor of a dreamlike field
- − Slightly less 'anime' in line work compared to typical Ghibli style
HiDream I1 Full
- + Strong Studio Ghibli character design aesthetic
- + Retains the urban street setting seen in the original image
- + Vibrant and clean illustration style
- − Fails to maintain the specific 'distracted' pose and looking-back dynamic
- − Introduces significant changes to the subjects' appearances and clothing
- − Mishandles the girl on the right, making her look too young and detached from the interaction
Verdict: FLUX.2 [dev] Flash is the clear winner because it successfully transforms the image into the requested art style while preserving the essential composition, poses, and narrative of the source meme. HiDream I1 Full produces a nice Ghibli-style illustration but loses the core 'distracted' dynamic of the original source photo.
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
FLUX.2 [dev] Flash
- + Excellent source preservation, maintaining the woman's identity and the dog exactly from the original.
- + Precise application of the requested edits with hair blowing and flying leaves.
- + High resolution and clarity consistent with the source image.
HiDream I1 Full
- + Successfully captures an 'energetic and lively feel' through a more dynamic pose.
- + Good interpretation of flying leaves with motion blur.
- − Very poor source preservation; the woman's face and clothes have changed, and the dog is completely removed.
- − The background and overall composition have been significantly altered from the source image.
Verdict: FLUX.2 [dev] Flash is the clear winner as it successfully applies the edit (hair movement and leaves) while perfectly preserving the person, the dog, and the background of the source image. HiDream I1 Full fails as an image editor, essentially generating a new image that removes the dog and changes the woman's appearance entirely.
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
FLUX.2 [dev] Flash
- + Excellent typography rendering including the accent on 'Caffè'.
- + Perfectly captures the requested 'subtle texture' and 'vintage' aesthetic.
- + Stronger vector emblem composition with a well-integrated banner.
- − The steam is a bit simplistic compared to the rest of the detailed illustration.
HiDream I1 Full
- + Clean, minimalist vector style.
- + Creative inclusion of a coffee cup icon inside the cloche.
- − Completely failed to include the main restaurant name 'Caffè Florian'.
- − The banner at the bottom is poorly rendered with jagged, asymmetrical edges.
- − Lacks the requested 'subtle texture' on the background.
Verdict: FLUX.2 [dev] Flash followed all instructions perfectly, including complex text rendering with proper accents and a beautiful vintage finish. In contrast, HiDream I1 Full failed to include the primary name of the restaurant and produced a lower-quality banner execution, making FLUX.2 for Dev the clear winner for this design task.
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
FLUX.2 [dev] Flash
- + Successfully includes all 6 requested steps of the mission.
- + Excellent text rendering for most labels including the astronaut names.
- + Captures a complex infographic layout with high detail.
- − Has some garbled text artifacts under the 'Launch' and 'Moon' labels.
- − Visual style is slightly more 'illustrative' than the requested 'flat-vector' style.
HiDream I1 Full
- + Perfectly adheres to the narrow flat-vector style with clean geometric shapes.
- + Solid color palette implementation according to the NASA-inspired request.
- + Very clean and modern aesthetic with crisp lines.
- − Fails to include all 6 requested steps, stopping halfway through the process.
- − Literally includes prompt text like '+ ICON' and 'UNLY' (presumably from only) in the labels.
- − Contains several gibberish text snippets instead of clear information.
Verdict: FLUX.2 [dev] Flash is the clear winner because it actually follows the multi-step instructions and renders most of the text correctly, providing a comprehensive mission overview. While HiDream I1 Full has a cleaner flat-vector aesthetic, it fails to generate the full sequence and includes prompt-engineering keywords as visible text on the poster.
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