OpenAI's previous generation image model with higher quality than DALL-E 2 and support for larger resolutions
Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.
DALL-E 3
18.5
arena score
#35 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]
24.5
arena score
#17 of 44 in Text-to-Image
Vote tally
Where the votes landed
DALL-E 3
0.0%
win rate
Ties
0.0%
FLUX.2 [dev]
100.0%
win rate
0.0%
0.0% ties
100.0%
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.”
DALL-E 3
FLUX.2 [dev]
0% wins
0% ties
100% wins
AI Judge Analysis
DALL-E 3
- + High artistic detail in the wood and lighting effects
- + Crisp focus and vibrant colors
- − Failed the spatial instructions significantly
- − Placed the book inside the cube and the sphere on top of the book
- − The cube has a thick wooden frame not mentioned in the prompt
FLUX.2 [dev]
- + Perfect adherence to all spatial instructions in the prompt
- + Engineered realistic glass reflections and refractivity
- + Accurately represents the soft window light from the left
- − Image is slightly less sharp compared to Model A
- − Composition is somewhat plain
Verdict: FLUX.2 [dev] followed every spatial instruction perfectly, placing the sphere inside the cube and the book on top. DALL-E 3 failed the physics and placement logic of the prompt, resulting in a visually pleasing but incorrect arrangement where the book is inside and the spheres/cube are framed in wood.
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.”
DALL-E 3
FLUX.2 [dev]
AI Judge Analysis
DALL-E 3
- + Excellent composition with a unique foreground framing and strong reflection in the puddle.
- + Atmospheric lighting that creates a cinematic mood despite the prompt's request for no stylization.
- + Captures the environment well with the lantern and traditional street feel.
- − Anatomical errors in the man's feet and crouched position.
- − Background vehicle lacks the requested motion blur, appearing mostly static.
- − The man's skin and proportions look somewhat illustrative and 'smooth' rather than natural.
FLUX.2 [dev]
- + Successfully incorporates the requested motion blur from the passing car.
- + Very realistic skin texture and hands, matching the 'natural skin' and 'candid' prompt requirements.
- + Accurate and complex mechanical detail of the red bicycle being repaired.
- − The framing is fairly standard and safe, lacking the 'imperfect' or highly creative edge.
- − Background reflections are less prominent than in the competing image.
Verdict: FLUX.2 [dev] followed the technical requirements of the prompt much better, specifically the 'motion blur' and 'natural skin texture' which DALL-E 3 struggled with. While DALL-E 3 provided a very artistic and interesting composition with the reflection, FLUX.2 [dev] achieved a much higher level of realism and better adherence to the specific camera characteristics requested.
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.”
DALL-E 3
FLUX.2 [dev]
AI Judge Analysis
DALL-E 3
- + Excellent depiction of intricate, weathered engraving on the plate armor
- + Stunning lighting with warm highlights and vibrant bokeh effects
- + Highly detailed, lifelike iris and eye textures
- − Hair braids feel secondary to the helmet and are less prominent
- − The composition is a bit tight, losing some of the promised leather and cloth detail
FLUX.2 [dev]
- + Perfect adherence to the 'hair braided with small beads' requirement
- + Skin texture and battle-worn scars look very realistic and natural
- + Clear detail on the leather straps and the cloth cowl as requested
- − Lighting is somewhat flat compared to the dramatic torchlight requested
- − Armor engraving is present but lacks the depth and craftsmanship of the competitor
Verdict: DALL-E 3 creates a more cinematic and artistically striking image with superior metal textures and lighting, while FLUX.2 [dev] provides better adherence to specific character details like the beaded braids and leather straps. FLUX.2 [dev] feels more like a raw photograph of a character, whereas DALL-E 3 feels like a high-end digital illustration.
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.”
DALL-E 3
FLUX.2 [dev]
AI Judge Analysis
DALL-E 3
- + Excellent depiction of god rays and dynamic lighting.
- + Captures a playful, high-energy 'tumbling' interaction.
- + Strong vibrant colors and wholesome atmosphere.
- − Has a very stylized, 3D-render look rather than 'hyper-photorealistic'.
- − Strange 'creature' hybrids where butterflies have fuzzy mammal heads.
- − Fox paws look more like cartoon mittens than realistic paws.
FLUX.2 [dev]
- + Successfully achieves a 'hyper-photorealistic' look with natural textures.
- + Includes all four requested animals with very accurate, realistic anatomy.
- + Beautifully rendered dew sparkles and natural sunrise lighting.
- − The animals are largely sitting and staring rather than 'tumbling together'.
- − The composition is a bit more static compared to the requested action.
Verdict: While DALL-E 3 captures the energy and magical lighting of the prompt, it fails the 'hyper-photorealistic' requirement and produces uncanny animal-headed butterflies. FLUX.2 [dev] provides a stunningly realistic image with accurate animal details and beautiful textures, making it the superior interpretation of the technical requirements.
Next steps
Explore each model
Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment