Black Forest Labs' flagship image generation model delivering state-of-the-art quality with exceptional realism, precision, and consistency for both text-to-image and advanced image editing
Settled by community votes across 7 shared challenges, with an AI judge weighing in on each.
FLUX.2 [max]
25.9
arena score
#10 of 62 in Text-to-Image
Skill signature
· Text-to-Image
P-Image Edit
9.5
arena score
#29 of 32 in Image Editing
Vote tally
Where the votes landed
FLUX.2 [max]
0%
win rate
Ties
0%
P-Image Edit
0%
win rate
Shared challenges
7
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Man and Car in California
EditingEdit instruction
“Make a photo of the man driving the car down the California coastline”
Source
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent preservation of the car's identity and details
- + Accurate recreation of the man's hair and clothing (plaid coat and scarf)
- + Realistic lighting and shadow integration
- − The man appears slightly too large relative to the car's interior
P-Image Edit
- + Beautiful landscape and coastal composition with motion blur
- + Successfully places the man and car in the requested environment
- − The man's clothing and facial details from the source image are significantly lost or obscured
- − The car's wheels exhibit motion blur artifacts that don't match the stationary body perfectly
- − The driver is far back in the seat and low in the frame
Verdict: Both models succeeded in the core task, but FLUX.2 [max] is the superior edit because it preserved specific details from both source images, including the man's distinct plaid coat and scarf. P-Image Edit created a more dynamic action shot with motion blur and palm trees, but it lost the visual identity of the subject in the process.
Pose & Character Mashup
EditingEdit instruction
“Use Image 1 as the exact pose reference and Image 2 as the character reference. Recreate the person/character from Image 2 in the exact dynamic pose and body position from Image 1. Keep the exact face, hair, clothing style/details, and expression from Image 2. Match the lighting and environment of Image 1. The final image must show the character from Image 2 performing the precise action/pose from Image 1 with perfect anatomy and natural integration.”
Source
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent character preservation of the face, sunglasses, and clothing from Image 2.
- + Recreates the lighting and yellow background of Image 1 perfectly.
- + Anatomy of the feet and torso is realistic and grounded.
- − Modifies the pose slightly from the source, making it less dynamic and more like a crouch.
- − The upper right hand has six fingers and a distorted structure.
P-Image Edit
- + Successfully copies the exact complex leg and arm positioning from Image 1.
- + Includes specific character elements like the sunglasses and scarf.
- − Major anatomically failures, including a foot where a hand should be and blended skin/clothing textures.
- − Unnaturally long hair and red sleeve remnants that do not match the character reference image.
- − The face and head shape significantly vary from the reference character.
Verdict: FLUX.2 [max] is the winner because it successfully creates a coherent, realistic person that looks exactly like the character in Image 2, despite simplifying the pose slightly. P-Image Edit attempts a more accurate recreation of the dynamic pose from Image 1, but the result is full of severe anatomical errors, including a foot growing out of an arm and mismatched limbs.
Outfit Transfer Challenge
EditingEdit instruction
“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”
Source
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Successfully replicates the specific plaid pattern and coat style from Image 2.
- + Maintains excellent facial fidelity and skin texture from Image 1.
- + Integrates small accessories like the watch and ring with high realism.
- − Crops the person to a medium-close shot, losing the requested shoe detail from the full outfit.
P-Image Edit
- + Attempts to include the full body and shoes as implied by the outfit request.
- + Preserves the overall color grade and background of the original image well.
- − Adds a large gold chain that was not present in Image 2.
- − Poorly integrates the face, creating a noticeable seam and 'pasted-on' appearance.
- − Alters the body proportions incorrectly, especially in the legs and hands.
Verdict: FLUX.2 [max] is the superior model because it maintains the high photographic quality of the source while accurately replicating the complex layers and patterns of the outfit. P-Image Edit fails significantly on technical execution, with poor mask blending on the face and the addition of accessories not found in the reference images.
Bald man challenge
Image EditingEdit instruction
“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent integration with original facial features
- + Hair texture matches the existing beard perfectly
- + Natural lighting and shadows on the forehead and hair
- − None notable
P-Image Edit
- + Successfully added a large volume of hair
- + Maintains original facial structure
- − Hair texture/curl style feels mismatched with the existing beard
- − The hairline integration is slightly less seamless than Model A
- − The curls have a repetitive, 'rendered' look in some areas
Verdict: FLUX.2 [max] provided a much more realistic and cohesive result, making the hair look like a natural extension of the person by matching the color and texture of the existing beard. P-Image Edit achieved the goal of adding hair, but the curly texture felt artificial and didn't blend as naturally with the original image's lighting and aesthetic.
Over-the-top cartoon caricature
EditingEdit instruction
“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”
Source
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent caricature style with exaggerated proportions.
- + Clearly incorporates all three elements: TV anchor, dogs, and hockey.
- + Maintains character likeness while translating it into a cartoon.
- − The text on the desk and scoreboard is nonsensical.
- − The character's eyes were changed from brown to blue.
P-Image Edit
- + Strong environment design featuring an arena with a crowd.
- + Good facial likeness and preservation of original eye color.
- + Successfully incorporates TV, hockey, and dog elements.
- − The 'dog' elements are small and oddly designed, particularly the red structure on the dog.
- − The hockey players in the background have distorted faces.
- − The text 'TV SHOW' is somewhat generic compared to a news-style logo.
Verdict: FLUX.2 [max] creates a much better caricature with more appeal and a cleaner art style, perfectly balancing the hockey sticks, numerous dogs, and the anchor desk. While P-Image Edit does a better job of preserving the subject's brown eyes and providing a grander arena setting, its execution of the dogs and background characters is distorted and less effective as a 'humorous caricature.'
Studio Ghibli Anime Style
EditingEdit instruction
“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”
Source
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent adherence to the specific Studio Ghibli art style including facial features and watercolor textures.
- + Maintains the composition and blurred background elements from the source image accurately.
- + Captures the warm, nostalgic, and dreamy atmosphere requested in the prompt.
- − The color palette is slightly washed out compared to some more vibrant Ghibli films.
P-Image Edit
- + Successfully converts the scene into an illustration while keeping the character likenesses high.
- + Incorporates a whimsical, dreamy background with stylized clouds and trees.
- − The art style feels more like modern generic anime or a digital manhwa rather than the specific Studio Ghibli style.
- − The character lighting is a bit flat compared to the requested 'gentle lighting' of Ghibli.
Verdict: FLUX.2 [max] significantly outperformed P-Image Edit by accurately capturing the specific hand-painted, watercolor aesthetic synonymous with Studio Ghibli. While P-Image Edit created a clean illustration, it leaned toward a generic vector art style and failed to replicate the iconic Ghibli facial structures and textures that FLUX.2 [max] mastered.
Golden Hour Stroll
Image EditingEdit instruction
“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
FLUX.2 [max]
P-Image Edit
AI Judge Analysis
FLUX.2 [max]
- + Excellent wind-blown hair effect that looks natural and directional.
- + The leaves have realistic depth and motion blur.
- + Very strong source preservation of the face, clothing, and dog.
- − Some leaves in the foreground are slightly oversized.
P-Image Edit
- + Adds a large amount of swirling hair for a high-energy look.
- + Successfully adds multiple leaves throughout the scene.
- + Maintains the likeness of the subject and dog well.
- − The hair edit creates some messy artifacts and looks less naturally integrated.
- − Leaves look a bit more like static icons rather than objects in motion.
- − The dog's tail is slightly altered in a way that looks less natural.
Verdict: FLUX.2 [max] is the winner because its edits feel more realistic and integrated into the scene. The wind effect on the hair is convincing and matches the direction of the flying leaves, whereas P-Image Edit's hair looks a bit more digitally manipulated and messy. FLUX.2 [max] also uses motion blur on the leaves to better convey the requested 'dynamic' feel.
Next steps
Explore each model
PrunaAI's sub-1-second multi-image editing model supporting up to 5 reference images with state-of-the-art quality