Black Forest Labs' open-weights image generation model with frontier performance, available for non-commercial local deployment
Settled by community votes across 3 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.
Seedream 4.0
24.6
arena score
#16 of 44 in Text-to-Image
Vote tally
Where the votes landed
FLUX.2 [dev]
0.0%
win rate
Ties
0.0%
Seedream 4.0
100.0%
win rate
0.0%
0.0% ties
100.0%
Shared challenges
3
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]
Seedream 4.0
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [dev]
- + Excellent handling of glass thickness and optical refraction.
- + The blue sphere is correctly placed inside the volume of the cube.
- + Soft lighting feels natural and consistent across the table and glass surfaces.
- − The sphere appears to be floating slightly rather than sitting on the bottom.
- − The plant in the background is very blurry, making it harder to see through the glass as requested.
Seedream 4.0
- + The cube geometry is very sharp and precise.
- + Reflections on the mirrored base of the cube add a nice level of detail.
- + The plant is clearly visible through the glass panels.
- − The red book is not perfectly centered and looks a bit small for the cube.
- − The perspective of the table edges is slightly inconsistent with the cube's orientation.
Verdict: Both models followed the complex spatial instructions perfectly. FLUX.2 [dev] produced a more photorealistic image with superior light and material physics, while Seedream 4.0 provided a cleaner, more literal interpretation of the 'visible through the glass' instruction. FLUX.2 [dev] is the winner for its more convincing textures and natural integration of the objects.
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]
Seedream 4.0
0% wins
0% ties
100% wins
AI Judge Analysis
FLUX.2 [dev]
- + Excellent depiction of motion blur with multiple passing cars that look realistic
- + Highly detailed skin textures and lifelike facial features
- + Very authentic rain effects and complex reflections on the asphalt
- − The bicycle frame geometry is slightly anatomically incorrect near the handlebars
- − The man's right leg appears to be missing or merged into the bike frame
Seedream 4.0
- + Captures a more traditional Japanese street vibe with tool details on the ground
- + The bicycle structure is more coherent and recognizable
- + Good use of puddle reflections and composition
- − Skin and hair textures look slightly smoothed and less like a 'natural' 50mm photo
- − The motion blur on the car is a bit muddy and looks like a digital filter rather than long exposure
- − Rain is barely visible compared to the request for light rain
Verdict: FLUX.2 [dev] significantly outperforms Seedream 4.0 in terms of photorealism, particularly in the rendering of the man's skin and the sophisticated motion blur of the traffic. While Seedream 4.0 has a more structurally sound bicycle, FLUX.2 [dev] captures the cinematic, high-fidelity aesthetic and the specific 'light rain' atmosphere requested in the prompt more effectively.
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]
Seedream 4.0
AI Judge Analysis
FLUX.2 [dev]
- + Excellent anatomical realism for all four animal types.
- + Subtle, natural lighting with very clean fur textures.
- + High level of detail in both the foreground subjects and the background meadow.
- − The scene is static and lacks the 'playfully chasing' and 'tumbling' action requested.
- − Includes an extra bunny not specified in the prompt.
Seedream 4.0
- + Captures the 'playfully chasing' and 'tumbling' action much more effectively.
- + Dynamic composition with animals in active poses.
- + Beautiful use of dew sparkles and directional god rays.
- − The kitten has a slightly distorted, less realistic face compared to the others.
- − The fox's red fur is a bit oversaturated, looking less photorealistic than Image A.
Verdict: While FLUX.2 [dev] produces much higher anatomical realism and cleaner textures, it fails to capture the dynamic action of the prompt, showing the animals sitting still. Seedream 4.0 much better reflects the requested energy of animals 'tumbling' and 'chasing,' despite having slightly more 'AI-like' facial features and saturated colors.
Next steps
Explore each model
ByteDance's image generation model with integrated text-to-image and image editing capabilities in a unified architecture, supporting up to 4K resolution