Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment
Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.
FLUX.2 [dev]
24.5
arena score
#17 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.
FLUX.2 [dev] Flash
26.8
arena score
#5 of 44 in Text-to-Image
Vote tally
Where the votes landed
FLUX.2 [dev]
0.0%
win rate
Ties
33.3%
FLUX.2 [dev] Flash
66.7%
win rate
0.0%
33.3% ties
66.7%
Shared challenges
4
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]
FLUX.2 [dev] Flash
0% wins
100% ties
0% wins
AI Judge Analysis
FLUX.2 [dev]
- + Excellent rendering of the blue sphere with realistic internal light refraction.
- + Soft, warm lighting that convincingly emanates from the left window.
- + Smooth glass texture with realistic rounded edges.
- − The plant in the background is very blurry and lacks definition.
- − The blue sphere appears to be magically floating rather than resting on the bottom of the cube.
FLUX.2 [dev] Flash
- + Superior sharpness and detail in the plant leaves and the book's spine text.
- + The glass cube features realistic dust particles and reflections that ground it in the environment.
- + Perfect adherence to the spatial arrangement, with the sphere correctly resting on the glass floor.
- − The lighting is slightly more clinical and less 'soft' than requested compared to Model A.
Verdict: Both models followed the complex prompt perfectly, including the specific spatial relationships of all four objects. FLUX.2 [dev] Flash is the winner because it provides much higher detail in the secondary elements like the plant and book spine, and realistically places the sphere on the floor of the cube, whereas FLUX.2 [dev] shows the sphere floating.
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]
FLUX.2 [dev] Flash
AI Judge Analysis
FLUX.2 [dev]
- + Excellent motion blur on passing vehicles
- + Authentic 'imperfect framing' with a candid street photography feel
- + Subtle, realistic rain effects
- − The structural integrity of the bicycle's handlebars and frame is messy
- − Skin texture is slightly smoothed compared to the prompt's request
FLUX.2 [dev] Flash
- + Very high detail in skin texture and facial features
- + Better bicycle coherence with a clear frame and basket
- + Stronger puddles and rain splashes on the ground
- − Motion blur on cars is less convincing than Model A
- − The composition feels slightly more staged than Model A
Verdict: Both models followed the prompt well, but Model A (FLUX.2 [dev]) captures the 'candid' and 'motion blur' requirements more effectively, creating a more cinematic atmosphere. However, Model B (FLUX.2 [dev] Flash) provides much better anatomical and mechanical detail, with superior skin textures and a more realistically constructed bicycle.
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]
FLUX.2 [dev] Flash
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [dev]
- + Excellent warm lighting with strong rim light on the character
- + Highly realistic leather texture and metallic reflections
- + More organic and believable integration of braids and beads
- − The scar on the forehead looks slightly like a surface-level sticker compared to the skin texture
FLUX.2 [dev] Flash
- + Very intricate engraving details on the plate armor
- + Symmetrical composition with balanced torchlight
- + Better representation of realistic skin trauma and dirt
- − Slightly flatter lighting compared to the dynamic depth of the other model
- − Metal textures look a bit cleaner/newer despite the battle-worn prompt
Verdict: Both models followed the prompt exceptionally well, capturing the braided hair, ornate armor, and lighting. FLUX.2 [dev] produces a more cinematic shot with superior lighting and depth, whereas FLUX.2 [dev] Flash provides more intricate fine-line engravings on the armor and a more symmetrical composition.
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]
FLUX.2 [dev] Flash
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [dev]
- + Excellent photorealistic lighting with realistic god rays and backlighting on the fur.
- + Very high-quality textures on the animals, particularly the fox and golden retriever.
- + The animals' eye expressions are sharp and soulful.
- − Includes an extra baby bunny not explicitly requested (two instead of one).
- − The kitten appears somewhat static and less 'playful' compared to the others.
FLUX.2 [dev] Flash
- + Better captures the 'playfully chasing' and 'tumbling' aspect of the prompt with dynamic poses.
- + Accurately includes more butterflies scattered throughout the scene.
- + Bright, vibrant colors that enhance the wholesome vibe.
- − Anatomy errors: the kitten has an extra leg/paw under its chest, and the brown bunny has facial features that look like a cat/bunny hybrid.
- − The lighting feels slightly more filtered/digital and less naturally photorealistic than Model A.
Verdict: FLUX.2 [dev] produces a much more polished and photorealistic image with superior fur textures and realistic lighting, though it added an extra rabbit. FLUX.2 [dev] Flash better captured the playful movement requested in the prompt, but it suffered from notable anatomical glitches and a more artificial aesthetic. FLUX.2 [dev] is the winner for its overall technical quality and believable rendering.
Next steps
Explore each model
Fast distilled version of Black Forest Labs' FLUX.2 [dev] optimized for speed and cost efficiency.