Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment
Settled by community votes across 13 shared challenges, with an AI judge weighing in on each.
FLUX.2 [dev]
#18 of 62 in Text-to-Image
Qwen Image
#35 of 62 in Text-to-Image
Where the votes landed
FLUX.2 [dev]
0%
win rate
Ties
0%
Qwen Image
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]
- + Excellent realistic textures on the wooden table and book cover.
- + Accurate glass physics including refraction and subtle dust/imperfections.
- + Perfect adherence to spatial instructions with the plant clearly visible behind the cube.
- − The sphere's reflection on the bottom glass seems slightly offset.
Qwen Image
- + Clean, minimalist composition.
- + Distinct rendering of the requested elements (red book, blue sphere, green plant).
- − Internal reflections in the glass cube are physically confusing and busy.
- − The sphere appears to be floating rather than resting on the bottom surface.
- − The scale of the sphere is slightly larger, making it look less like the requested 'small' sphere.
Verdict: FLUX.2 [dev] produces a significantly more realistic and physically coherent image, with superior material textures and natural light interactions through the glass. Qwen Image achieves the prompt requirements but suffers from 'AI-look' artifacts, particularly in how the sphere sits within the cube and the unnatural internal reflections.
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]
- + Excellent realism with highly detailed skin textures and wet hair.
- + Captures motion blur from background traffic flawlessly as requested.
- + Superior lighting and reflections on the pavement and jacket.
- − The bicycle geometry is slightly confusing near the handlebars with overlapping frames.
- − The man's fingers exhibit some minor anatomical distortion while gripping the tools.
Qwen Image
- + Shows the full body and bicycle, making the 'repair' action very clear.
- + Good adherence to the 'reflections on wet pavement' requirement.
- − Failed to produce the requested motion blur on the background cars.
- − The rain effect looks like static vertical brush strokes rather than a natural atmosphere.
- − The sidewalk and bicycle stand are visually glitchy, appearing to melt into the ground.
Verdict: FLUX.2 [dev] significantly outperforms Qwen Image in terms of photorealism and adherence to complex camera instructions like motion blur. While Qwen Image provided a better full-body composition, it suffered from severe technical artifacts and failed to render the requested rain and motion effects convincingly.
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]
- + Excellent adherence to the 'beads in braids' prompt with natural placement
- + Extremely realistic skin texture and lifelike eyes
- + Detailed engraving on the armor that feels physically plausible
- − The torch in the background is slightly soft and lacks detail
- − The sparks/bokeh are a bit uniform across the image
Qwen Image
- + Strong cinematic lighting with high contrast
- + Very sharp and detailed textures on the leather straps and chainmail
- + Dramatic facial expression that fits the 'battle-worn' theme
- − The beads appear to be floating or awkwardly attached to the top of the hair
- − The 'bokeh sparks' look like digital starburst brushes rather than realistic light particles
- − The armor engraving looks slightly more flat/geometric compared to the relief in A
Verdict: FLUX.2 [dev] produces a more grounded and realistic portrait with superior integration of the hair beads and skin textures. While Qwen Image offers a more dramatic, high-contrast look and excellent chainmail detail, it fails on the 'bokeh' effect and has less natural-looking hair accessories.
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]
- + Excellent typography with a professional layout resembling a real menu.
- + High-quality, realistic food photography.
- + Includes distinct, well-organized sections for Appetizers, Pizza, and Mains.
- − Nonsensical text and gibberish titles like 'PIZZAU'.
- − Text alignment becomes messy in the price columns.
Qwen Image
- + Creative grid layout with bold, saturated accent colors behind food items.
- + Legible header fonts and a clean, minimalist aesthetic.
- + Better adherence to the 'modern minimalist' style requested.
- − Food photos appear slightly AI-generated and plastic rather than realistic.
- − Consolidates categories into 'Pizza/mains' rather than having three distinct sections.
- − Prices are overly simplified to single digits/unrealistic numbers.
Verdict: FLUX.2 [dev] produces a more traditional and realistic menu layout that looks ready for a casual dining table, though it struggles with consistent spelling. Qwen Image offers a more modern, graphically bold interpretation that captures the 'minimalist' and 'vibrant accents' part of the prompt more effectively, despite the food quality being lower. Qwen Image is the winner for better capturing the specific aesthetic vibe and layout requested in the prompt.
Magic Burger Explosion: Fiery Photorealism Challenge
Text-to-Image“Ad for 'Magic Burger'. Dynamic, exploded burger with all components (bun, patty, cheese, lettuce, tomato, sauce) suspended in mid-air. Emphasize photorealistic detail and a sense of motion. Dark, fiery background with glowing embers. Integrate text: 'MAGIC BURGER' as a prominent title, 'LIMITED TIME ONLY' as a secondary message, and '€6.99' in a starburst, all rendered with a fiery, glowing effect.”
AI Judge Analysis
FLUX.2 [dev]
- + Excellent text integration with a consistent fiery glow effect across all elements.
- + Photorealistic textures on the burger bun and patty.
- + Clean composition that follows the 'exploded' instructions clearly.
- − The 'exploded' effect is a bit static, with components just stacked vertically rather than dynamically scattered.
Qwen Image
- + High sense of motion with ingredients flying at different angles.
- + Very vibrant fiery background with clear glowing embers and debris.
- + Includes extra details like sesame seeds and onions not explicitly requested but adding to the appeal.
- − The 'LIMITED TIME ONLY' text lacks the requested fiery glowing effect found on other elements.
- − The starburst element looks a bit more like a flat sticker compared to the integrated glow of Model A.
Verdict: FLUX.2 [dev] followed the text styling instructions perfectly, ensuring every piece of text had the requested fiery glow. However, Qwen Image created a much more dynamic and energetic 'exploded' burger composition that felt more like a professional advertisement, despite missing the glow effect on one line of text.
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]
- + Excellent chalk texture with realistic smudging and dusty grain
- + Highly accurate text rendering with correct spelling for all items
- + Natural handwriting variation that matches the 'chalkboard' brief perfectly
- − The layout for the second item is slightly inconsistent with the others
Qwen Image
- + Clean, legible layout with consistent spacing
- + Warm, inviting cafe background composition
- − Serious factual error in the date rendering it as '20026' instead of '2026'
- − Spelling error in 'Risotto' (spelled as 'Risoto')
- − Text looks more like a digital font or vector stroke than real chalk material
Verdict: FLUX.2 [dev] significantly outperforms Qwen Image by adhering to the complex prompt's request for a realistic chalk texture and accurate handwriting. While Qwen Image provides a nice composition, it fails on basic text accuracy, misspelling 'Risotto' and incorrectly rendering the year as 20026.
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]
- + Excellent texture on the astronaut suit and horse hair.
- + More realistic lighting and cinematic color palette.
- + Superior background detail with a vibrant nebula and realistic Earth edge.
- − Failed the logic trap of 'horse on top' of the astronaut.
Qwen Image
- + Clean, well-defined subjects.
- + Good anatomical consistency for the horse.
- + Consistent composition with the astronaut riding the horse.
- − Failed the logic trap of 'horse on top' of the astronaut.
- − Lighting feels flat and less cinematic compared to FLUX.2.
- − Lower level of detail in the background and surface textures.
Verdict: Both models failed the specific logic trap included in the prompt ('horse on top, not vice versa'), instead generating a standard astronaut-riding-horse image. FLUX.2 [dev] is the superior image due to its much higher fidelity, realistic lighting, and detailed textures on both the horse and the space suit, whereas Qwen Image looks more like a standard 3D render.
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]
- + Excellent adherence to the 'inside' perspective requested in the prompt
- + Superior rendering of capybara fur texture and paws
- + Captures the bored expression of the high-detail background passenger perfectly
- − The capybara's head is slightly oversized for the car cabin
Qwen Image
- + Good depth of field and nighttime bokeh effect through the windows
- + Includes a taxi sign on the roof which adds to the setting
- + Clean and smooth aesthetic
- − The capybara's hands look more like primate hands than capybara paws
- − The passenger is looking away from her phone rather than at it as requested
- − Composition feels less like being 'inside' the taxi and more like looking through a door
Verdict: FLUX.2 [dev] provided a more accurate interpretation of the prompt, particularly with the passenger's behavior and the capybara's anatomy. Qwen Image produced a high-quality visual, but failed to show the passenger actually looking at her phone and gave the capybara human-like fingers.
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]
- + Perfect text rendering with no spelling errors.
- + Excellent integration of all design elements like the thorny border and parchment edges.
- + High-quality cinematic lighting around the central jack-o-lantern.
- − The twisted trees in the background are somewhat obscured by the dark fog compared to others.
Qwen Image
- + Strong atmospheric depth with clear silhouettes of twisted trees.
- + Good parchment texture and border design.
- − Several spelling errors in the main title including 'Halle Party'.
- − Text at the bottom uses a generic sans-serif font that clashing with the gothic theme.
- − Layout is a bit cluttered with overlapping bats and text elements.
Verdict: FLUX.2 [dev] followed the prompt instructions perfectly, producing professional-grade typography and a cohesive gothic aesthetic. In contrast, Qwen Image struggled with spelling the request title correctly and used an unfitting modern font for the event details, which broke the vintage theme.
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]
- + Excellent text rendering and alignment following the prompt instructions.
- + Highly detailed PBR textures on the fish and rice that give a high-end 3D render feel.
- + Perfect adherence to the isometric perspective and clean diorama base requirement.
- − The color of the word 'JAPAN' is dark grey/black instead of white, which slightly affects the aesthetic balance against the background.
Qwen Image
- + Strong 'cartoon' aesthetic with soft, rounded shapes.
- + Good use of color contrast in the typography against the light blue background.
- + High quality 3D lighting and soft shadows.
- − The flag icon in the text area is slightly distorted and contains an extra orange artifact below it.
- − The isometric perspective is slightly inconsistent, feeling more like a standard perspective shot.
- − The sushi rice looks somewhat like large plastic beads rather than realistic soft textures.
Verdict: FLUX.2 [dev] followed the prompt's layout instructions much more precisely, particularly in regards to the top-center text alignment and the realistic PBR material request. While Qwen Image created a charming cartoon scene, it struggled with the accuracy of the flag icon and the specific text placement requested. FLUX.2 [dev] is the winner for its superior clarity, technical execution of textures, and perfect adherence to the isometric diorama style.
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]
- + Excellent fur texture rendering and distinct lighting effects on the hair
- + Includes two bunnies which adds to the 'fluffy' feel of the scene
- + Realistic interactions between the animals and the environment
- − The animals are sitting relatively still rather than 'playfully chasing' or 'tumbling'
- − The kitten's eye placement is slightly off-kilter
Qwen Image
- + Successfully captures the 'playfully chasing' movement requested in the prompt
- + Superior 'god rays' effects that create a more dramatic sunrise atmosphere
- + Clearer, more appealing facial expressions on all animals
- − The front paws of the fox are poorly rendered and appear as dark, undefined blobs
- − Background butterflies appear somewhat flat compared to the foreground
Verdict: Qwen Image is the winner because it better captures the dynamic energy of 'playfully chasing and tumbling' requested in the prompt, whereas FLUX.2 [dev] depicts a static portrait. Although FLUX.2 [dev] has slightly more consistent paw anatomy, Qwen Image provides a more magical atmosphere with superior god rays and more expressive, joyful faces.
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]
- + Excellent typography with perfect spelling and accent marks.
- + Professional vector composition with a well-integrated banner.
- + Accurate adherence to the vintage minimalist aesthetic.
- − The banner lines under the text are slightly cluttered.
Qwen Image
- + Successfully captured the requested color palette and cloche theme.
- + Good rendering of the 'Est. 1720' text within the ribbon.
- − Major spelling and typographic errors in the main name.
- − Overlapping text elements create illegibility.
- − Less sophisticated vector style compared to the competitor.
Verdict: FLUX.2 [dev] produced a high-quality, professional logo that followed all prompt instructions perfectly, including complex text rendering with proper accents. Qwen Image struggled significantly with the typography, resulting in jumbled and misspelled text, which makes it unusable as a logo.
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]
- + Includes all nine requested icons and details in a structured grid.
- + Highly polished, professional vector aesthetic with consistent lighting.
- + Mostly accurate rendering of long-form text and names like 'Armstrong' and 'Collins'.
- − The layout is cluttered with redundant icons, creating some visual confusion.
- − Several labels have garbled text that is difficult to read.
- − Text labels like 'SATURN ICON' include the literal word 'icon' from the prompt instead of just the object name.
Qwen Image
- + Effective use of negative space and large typography.
- + Clear, bold iconography that matches the 'flat-vector' request well.
- + Sequential layout uses arrows to convey a path, which fits the mission-step theme.
- − Included meta-instructions '(Stop at landing)' as literal text in the image.
- − Significant spelling errors such as 'Sarth Orbit', 'Apoll', and 'Arnnstrong'.
- − Missed several discrete steps requested in the prompt.
Verdict: FLUX.2 [dev] provides a much more comprehensive set of icons that more closely follows the multi-step mission sequence, and its visual fidelity is significantly higher. Qwen Image fails by including internal prompt instructions like '(Stop at landing)' as part of the poster's title and suffers from multiple severe spelling errors and chronological gaps. Despite some garbled text at the bottom, FLUX.2 [dev] is the superior choice for a structured infographic.
Explore each model
Alibaba's Qwen image model