FP8 quantized variant of Black Forest Labs' FLUX.1 [schnell] model, offering ~2x faster inference with reduced precision while maintaining high-quality image generation in 4 steps
Settled by community votes across 12 shared challenges, with an AI judge weighing in on each.
FLUX.1 [schnell] FP8
#47 of 62 in Text-to-Image
HiDream I1 Full
#60 of 62 in Text-to-Image
Where the votes landed
FLUX.1 [schnell] FP8
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.1 [schnell] FP8
- + Excellent sharp resolution and clarity
- + Vibrant colors and realistic material textures
- + Accurate depiction of the glass cube and sphere
- − The glass cube has an illogical horizontal internal shelf
- − The cube looks more like a rectangular prism than a perfect cube
HiDream I1 Full
- + Better 'cube' proportions
- + Realistic natural window lighting and shadows
- + Highly accurate adherence to the plant being visible 'through' the glass
- − The blue sphere appears to be floating mid-air without support
- − Slightly softer focus compared to Model A
Verdict: Both models followed the prompt instructions very well, including the spatial relationships of the objects. FLUX.1 [schnell] FP8 produced a cleaner, sharper image with beautiful textures, but HiDream I1 Full captured the logic of the scene better despite the floating sphere, specifically regarding the plant's visibility through the glass and the natural lighting.
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.1 [schnell] FP8
- + Excellent reflection quality on the wet pavement.
- + Realistic bicycle anatomy and scale.
- + Good capture of shallow depth of field with cinematic lighting.
- − The man's ethnicity appears somewhat ambiguous or non-Japanese.
- − Missing the specific atmospheric 'rain' effect beyond just wet ground.
- − The bicycle handlebars and cables are a bit messy near the break levers.
HiDream I1 Full
- + Captures the 'light rain' atmosphere more effectively with visible droplets.
- + Natural skin texture and expressive facial features.
- + Stronger 'candid' feel with the subject sitting by the side of the road.
- − The red bicycle is disproportionately small, appearing like a child's bike.
- − Failed to include 'motion blur from passing cars' as the background vehicles are sharp.
- − Anatomy error regarding the left hand interacting with the seat.
Verdict: FLUX.1 [schnell] FP8 produced a more coherent composition with correct proportions for the bike and man, and better lighting. However, HiDream I1 Full excelled at the 'Japanese' ethnic descriptor and the rainy atmosphere, though it failed significantly on the scale of the bicycle and missed the motion blur requirement. FLUX.1 [schnell] is the winner for its superior technical realism and adherence to the car motion blur prompt.
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.1 [schnell] FP8
- + Strong cinematic lighting with convincing warm torchlight reflections
- + Hyper-realistic skin texture and lifelike eyes
- + Clean, high-resolution rendering with sophisticated color grading
- − Minimal braiding visible compared to the prompt requirements
- − The scars are extremely faint, bordering on non-existent
- − Armor engraving is partially obscured by the tight framing
HiDream I1 Full
- + Excellent adherence to specific details like braided hair with beads
- + Clearly visible scars and dirt, better representing the battle-worn theme
- + Highly detailed ornate engraving on the plate armor is well-displayed
- − Lighting feels slightly more artificial and flat than Model A
- − The face has a slightly more 'digital' or smoothed look compared to the realistic skin in A
- − Some minor artifacts in the background bokeh sparks
Verdict: While FLUX.1 [schnell] FP8 produces a more lifelike and photographically impressive portrait, HiDream I1 Full is the clear winner for prompt adherence. HiDream I1 Full successfully incorporated the complex braiding, beads, and battle-worn scars that were mostly lost or softened in the FLUX.1 output.
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.1 [schnell] FP8
- + Excellent structure that looks like a real multi-page menu.
- + Organized text sections with alignment and pricing consistency.
- + Higher quantity of food items which creates a more realistic grid layout.
- − Several spelling errors in headers (e.g., 'APPTIZERS', 'PIZZAL', 'SECCER').
- − The food images lack variety, appearing as circular tarts or pizzas in every section.
HiDream I1 Full
- + Clean and legible typography for the main headers.
- + High-quality food photography with realistic textures and lighting.
- + Better spelling in the header sections.
- − Layout is overly simplistic for a restaurant menu.
- − Lack of section diversity as every single photo is of the same/similar pizza.
- − Poor information density with large empty spaces and few menu items.
Verdict: FLUX.1 [schnell] FP8 provides a much more convincing menu layout with professional alignment and a comprehensive structure, though it suffers from typical AI spelling issues in the headers. HiDream I1 Full produces cleaner images and text but fails the layout challenge by presenting a very basic, repetitive design that lacks the 'grid' and 'sections' requested in a meaningful way. FLUX.1 is better for a design professional seeking a template structure.
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.1 [schnell] FP8
- + Excellent chalk texture and natural handwriting variation.
- + Followed the specific menu request with readable (though slightly garbled) words.
- + Accurate date rendering and believable chalkboard reflections.
- − Text contains several spelling errors and repetitive lines.
- − Did not use elegant cursive for the title as requested.
HiDream I1 Full
- + Successfully used elegant cursive for the title as requested.
- + Clean, high-quality chalkboard illustration with a nice coffee cup detail.
- − Failed almost entirely on prompt adherence for menu items.
- − The text is largely gibberish/placeholder characters.
- − The handwritten style looks more like a digital font than natural chalk.
Verdict: FLUX.1 [schnell] FP8 captured the requested menu items and realistic chalk texture much better, even though it struggled with some spelling. HiDream I1 Full followed the instruction for cursive titles but failed to include any of the specified food items, resulting in a board full of illegible gibberish.
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.1 [schnell] FP8
- + Excellent adherence to the 'horse on top' spatial instruction
- + Cinematic composition with high-quality rendering of the planet and lighting
- + Creative interpretation of equipment designed for a horse
- − The anatomy of the two distinct horse heads on one body is confusing and likely an artifact
- − The 'astronaut' part of the image is replaced by a box rather than a recognizable person
HiDream I1 Full
- + High visual clarity and classic cinematic aesthetic
- + Good rendering of the astronaut and horse anatomy
- − Completely failed the negative constraint to put the horse on top
- − Common interpretation that lacks the requested surrealism
Verdict: FLUX.1 [schnell] FP8 is the clear winner for its superior prompt adherence, successfully placing the horse on top of the equipment as requested, whereas HiDream I1 Full ignored the specific spatial instruction and reversed the subjects. While FLUX.1 suffered from some anatomical glitches involving multiple horse heads, it captured the 'surreal' and 'cinematic' aspects of the prompt far more effectively.
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.1 [schnell] FP8
- + Successfully renders clear 'TAXI' text on the cap
- + High quality texture on the capybara's fur
- + Effective blurred New York street background
- − The passenger is holding two phones simultaneously
- − The capybara's paw placement is somewhat messy on the wheel
- − The passenger is sitting in the front passenger seat instead of the back seat
HiDream I1 Full
- + Correctly places the woman in the back seat as requested
- + Excellent posture and paw placement for the capybara driver
- + Higher level of photorealism in the lighting and skin textures
- − Minor distortion on the text on the capybara's hat
- − The passenger's facial features are slightly blurry compared to the driver
Verdict: HiDream I1 Full is the clear winner as it followed the spatial instructions much better, correctly placing the woman in the back seat whereas FLUX.1 [schnell] FP8 placed her in the front. HiDream I1 Full also provided a more logical interaction with the steering wheel and avoided the 'double phone' artifact found in the FLUX image.
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.1 [schnell] FP8
- + Strong cinematic lighting with a vibrant glowing jack-o-lantern.
- + Captures a dark, moody gothic atmosphere effectively.
- + Good layout of the twisted trees and bats in the background.
- − Significant text rendering errors and typos throughout the invitation details.
- − The border appears more like ragged leather/fabric than a spooky thorn/web design.
HiDream I1 Full
- + Excellent adherence to the 'web and thorns' border request.
- + The scroll banner text 'You are invited to a night of frights' is rendered perfectly.
- + Clear hierarchy between the gothic title and the event details below.
- − The actual event details (date/time/location) are mostly nonsense or hallucinated strings.
- − The lighting is flatter and less cinematic compared to Model A.
Verdict: HiDream I1 Full followed the specific design elements of the prompt much better, including the webs and thorns border and the scroll banner text. While Flux.1 [schnell] FP8 had more impressive lighting and atmosphere, its text rendering and adherence to specific design details were significantly worse.
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.1 [schnell] FP8
- + Excellent 3D soft lighting and shadows.
- + High-quality material rendering for the rice and fish.
- + Correct implementation of the raised diorama base.
- − Failed the text request, outputting 'JAPAN JAPAN' instead of 'JAPAN SUSHI'.
- − Missing the small flag icon.
- − The text was somewhat cut off or poorly positioned at the top.
HiDream I1 Full
- + Perfect text rendering of 'JAPAN' and 'SUSHI'.
- + Clean isometric composition with varied sushi types.
- + Vibrant colors and high visual clarity.
- − Missing the requested small flag icon.
- − Background has subtle texture artifacts instead of being a solid color.
- − The base is a tray rather than a raised diorama pedestal.
Verdict: HiDream I1 Full is the clear winner because it correctly followed the complex text instructions and maintained a clean isometric style. FLUX.1 [schnell] FP8 produced superior textures and lighting, but failed significantly on the text rendering and text layout.
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.1 [schnell] FP8
- + Excellent depiction of warm golden light and god rays
- + Intricate fur texture and expressive facial features on the animals
- + Complex composition with many butterflies and a dense flower meadow
- − Failed to include a rabbit in the scene
- − Included extra animals that look like cat-fox hybrids, making the count five instead of four
- − Slightly less 'photorealistic' and more 'digital art' in style
HiDream I1 Full
- + Captures a more realistic photographic style
- + Accurately represents the tabby kitten requested in the prompt
- + Good spatial balance between the animals and the background light
- − Failed to include a rabbit in the scene
- − Anatomical oddity with the fox kit having a secondary tail-like shape or misaligned back leg
- − The butterflies appear a bit 'pasted on' rather than integrated into the lighting of the scene
Verdict: Both models failed to include the baby bunny from the prompt, showing a recurring difficulty with multi-subject animal scenes. FLUX.1 [schnell] FP8 produced a more magical, glow-filled image with higher detail in the fur, but it added extra hybrid animals. HiDream I1 Full was closer to the requested 'tabby' pattern for the kittens and felt more like a real photograph, despite some anatomical blurring on the fox.
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.1 [schnell] FP8
- + Features a high-quality vintage emblem aesthetic.
- + Correctly uses a cream and warm brown color palette with a textured background.
- − Failed significantly on the primary text rendering, outputting 'AFE FLAMILAN' instead of 'Caffè Florian'.
- − The architectural dome structure looks more like a building roof than a restaurant cloche dome.
HiDream I1 Full
- + Closer interpretation of a restaurant cloche dome with steam.
- + Clean, minimalist vector style that adheres well to the 'vector emblem' prompt.
- + Accurate rendering of the requested 'Est. 1720' banner.
- − Completely failed to include the brand name 'Caffè Florian' in the logo.
- − Composition is a bit generic and lacks the sophisticated typography requested.
Verdict: FLUX.1 [schnell] FP8 captures the vintage aesthetic and texture much better than its competitor, but it fails significantly on spelling, rendering 'AFE FLAMILAN' instead of the requested name. HiDream I1 Full provides a cleaner minimalist design and follows the 'cloche dome' prompt more accurately but omits the brand name entirely. HiDream I1 Full is the likely winner for better adhering to the specific icon requested and maintaining legible text, whereas FLUX suffers from a messy layout and severe typos.
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.1 [schnell] FP8
- + Stronger adherence to the chronological step-by-step layout requested.
- + Better use of the NASA-inspired color palette.
- + Includes more of the requested icon elements like the lunar module descent.
- − Text is largely gibberish except for headers.
- − Iconography is somewhat generic and abstract.
HiDream I1 Full
- + High-quality vector aesthetic with clean lines.
- + Very legible typography for the main labels.
- + Better visual balance for a poster format.
- − Fails to include all 6 requested steps, stopping effectively at step 3.
- − Icons are more decorative than instructional, missing the specific trajectory and landing steps.
Verdict: FLUX.1 [schnell] FP8 followed the instructional structure of the prompt more closely, attempting to visualize all six stages from launch to landing, though its small text is unreadable. HiDream I1 Full produced a more professional-looking poster with better typography and clean vector art, but it significantly failed on prompt adherence by omitting the lunar phases of the mission. FLUX.1 [schnell] FP8 is the winner for capturing the full sequence of the requested infographic.
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