Black Forest Labs' state-of-the-art image generation model with maximum quality and speed, supporting text-to-image and multi-reference image editing with up to 4MP output
Settled by community votes across 6 shared challenges, with an AI judge weighing in on each.
FLUX.2 [pro]
26.3
arena score
#9 of 44 in Text-to-Image
Skill signature
Not enough comparable category data
The chart appears once both models have ratings across at least three shared arena categories.
Qwen Image 2512
22.4
arena score
#26 of 44 in Text-to-Image
Vote tally
Where the votes landed
FLUX.2 [pro]
66.7%
win rate
Ties
16.7%
Qwen Image 2512
16.7%
win rate
66.7%
16.7% ties
16.7%
Shared challenges
6
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 [pro]
Qwen Image 2512
100% wins
0% ties
0% wins
AI Judge Analysis
FLUX.2 [pro]
- + Excellent photographic realism and lighting
- + Highly accurate refractions through the glass
- + Perfect adherence to all spatial instructions
- − The plant in the background is quite blurry due to shallow depth of field
Qwen Image 2512
- + Strong colors and clear visibility of the plant through the glass
- + Good texture on the book and wooden table
- − The glass cube has strange internal reflections that look like extra blue spheres
- − The 'cube' is not a perfect cube and lacks a clear top plane for the book to sit on
Verdict: FLUX.2 [pro] followed the prompt perfectly, delivering a high-quality photographic image with realistic glass refractions and consistent lighting. Qwen Image 2512 struggled with the geometry of the glass cube and generated confusing internal reflections that looked like additional blue spheres.
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 [pro]
Qwen Image 2512
100% wins
0% ties
0% wins
AI Judge Analysis
FLUX.2 [pro]
- + Excellent skin and hair textures that look highly realistic.
- + Strong adherence to the 'repairing' action mentioned in the prompt.
- + Superior technical composition with realistic reflections and water droplets on the bike.
- − The bike kickstand appears to be floating or disconnected from the frame.
Qwen Image 2512
- + Good bokeh and atmosphere that captures the 'street photo' aesthetic.
- + Captures the 'imperfect framing' well with the subject positioned centrally but slightly low.
- − The subject is posing/looking at the camera rather than repairing the bike.
- − Significant anatomical issues with the hands, notably the hand resting on the seat.
Verdict: FLUX.2 [pro] followed the prompt much more accurately, depicting the man in the act of repairing the bicycle with a candid feel, whereas Qwen Image 2512 felt more like a posed portrait. FLUX.2 [pro] also demonstrated significantly better detail in the skin, clothing, and mechanical parts of the bike, while Qwen Image 2512 struggled with distorted hand anatomy.
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 [pro]
Qwen Image 2512
50% wins
50% ties
0% wins
AI Judge Analysis
FLUX.2 [pro]
- + Excellent text legibility and font consistency.
- + Very clean, professional layout that matches the 'modern minimalist' prompt perfectly.
- + High-quality, realistic food photography that fits the grid layout well.
- − Spelling error in the 'MAINS' heading (rendered as 'MINS').
- − Repeated list items ('Garlic Bread' appears multiple times in different sections).
Qwen Image 2512
- + Stronger adherence to the 'colorful food photos in grid' requirement with a larger, more vibrant grid.
- + More visually dynamic composition with better use of color accents.
- − Text is largely nonsensical and poorly rendered compared to typical professional standards.
- − The layout feels slightly cramped with the large grid taking up the majority of the top half.
Verdict: FLUX.2 [pro] is the superior model for this task because it generates a highly professional and functional menu layout with legible text and clean design. While Qwen Image 2512 produces a more vibrant image grid, its text rendering is poor and the overall composition feels less like a real menu and more like an abstract graphic.
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 [pro]
Qwen Image 2512
AI Judge Analysis
FLUX.2 [pro]
- + Excellent side-profile composition that feels like a cinematic movie frame.
- + Very realistic lighting integration between the city lights and the car interior.
- + The capybara's leather jacket and hat are rendered with high material fidelity.
- − The capybara's hands look more human-like/primate-like than capybara paws.
- − The human passenger is slightly out of focus compared to the driver.
Qwen Image 2512
- + Perfectly captures the 'bored' expression requested for the passenger.
- + The capybara's face is front-and-center with high detail on the fur and whiskers.
- + Great adherence to the request for 'both front paws on the steering wheel' from a front-facing perspective.
- − The capybara's paws look somewhat like human hands with dark skin rather than rodent paws.
- − The internal car lighting is slightly flat compared to the dynamic reflections in Image A.
Verdict: Both models followed the prompt exceptionally well, capturing the surrealism of the scene with a high degree of photorealism. FLUX.2 [pro] is preferred for its superior cinematic composition and lighting, which feels more like a real photograph from a side window, while Qwen Image 2512 captures the specific 'bored' facial expression of the passenger more accurately.
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 [pro]
Qwen Image 2512
AI Judge Analysis
FLUX.2 [pro]
- + Captures the dynamic 'playfully chasing' and 'tumbling' action from the prompt.
- + Excellent backlight and god rays with beautiful dew drop details on the grass.
- + Very high level of realism in the fur texture and anatomy.
- − Failed to include the baby bunny requested in the prompt.
- − The kitten has an extra-long, almost lemur-like striped tail.
Qwen Image 2512
- + Successfully included all four requested animals (puppy, kitten, bunny, fox).
- + Strong, clear lighting that creates a warm, wholesome atmosphere.
- + High level of fur detail and clear, expressive eyes on all subjects.
- − The composition is a static group portrait rather than the 'chasing and tumbling' action requested.
- − The fox kit's face looks slightly more like an adult fox than a baby kit.
- − Anatomical blending where the bunny and kitten are tucked under the dog's paws.
Verdict: FLUX.2 [pro] created a much more cinematic and dynamic scene that perfectly captured the lighting and movement requested, but it failed to include the bunny. Qwen Image 2512 followed the subject list perfectly by including all four animals, but it opted for a static pose rather than the requested 'tumbling' action. FLUX.2 [pro] is arguably the higher quality image artistically, but Qwen Image 2512 wins on strict prompt adherence.
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 [pro]
Qwen Image 2512
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [pro]
- + Perfect adherence to the 'minimalist' and 'vector emblem style' descriptors.
- + Excellent, clean typography that reflects a classic, upscale aesthetic.
- + Very clean layout with a balanced circular composition.
- − The steam is a bit abstract compared to the rest of the illustration.
- − The 'Est. 1720' banner is slightly less flourish-heavy than some 'vintage' styles.
Qwen Image 2512
- + Highly detailed and illustrative 'vintage' style with an etched woodcut feel.
- + Very dynamic steam elements and a beautifully rendered banner.
- + Rich, warm brown tones with a high-quality paper texture.
- − Fails the 'minimalist' part of the prompt by being very busy and ornate.
- − The script typography, while nice, is less legible as a 'classic' logo emblem than Model A's serif font.
- − The composition feels slightly bottom-heavy with the large banner.
Verdict: FLUX.2 [pro] followed the prompt more accurately by balancing the 'vintage' and 'minimalist' requirements to create a functional vector-style logo. While Qwen Image 2512 produced a beautiful, highly detailed illustration, it ignored the 'minimalist' constraint and resulted in an image that is more of a digital painting than a scalable emblem. FLUX.2 [pro] is the winner for superior typography and better adherence to the specified design style.
Next steps
Explore each model
Improved version of Alibaba's Qwen image model with better text rendering, finer natural textures, and more realistic human generation.