Black Forest Labs' precision image generation model with maximum control, reliable text rendering, and complete creative control supporting up to 4MP output
Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.
FLUX.2 [flex]
24.8
arena score
#14 of 62 in Text-to-Image
Skill signature
· Text-to-Image
HiDream E1
0.7
arena score
#33 of 32 in Image Editing
Vote tally
Where the votes landed
FLUX.2 [flex]
0%
win rate
Ties
0%
HiDream E1
0%
win rate
Shared challenges
4
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
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 [flex]
HiDream E1
AI Judge Analysis
FLUX.2 [flex]
- + Excellent preservation of the source image background and clothing
- + Highly realistic texture and natural hairline that blends perfectly with existing sideburns
- + Maintains facial features and lighting accurately
- − The hair length is relatively short for a 'full head of hair' request, though it fits the character well
HiDream E1
- + Provides a very full and thick head of hair as requested
- − Significantly alters the background landscape, turning a scrubland into a sandy desert
- − The hair texture looks artificial and 'plastic' with a visible halo artifact around the head
- − Distorts the person's face, adding excessive aging spots and changing the skin texture
Verdict: FLUX.2 [flex] is the clear winner as it successfully adds realistic hair while perfectly preserving every other detail of the source image. In contrast, HiDream E1 fails as an editing tool by hallucinating a new background, distorting the subject's face, and rendering hair that looks digitally painted rather than natural.
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 [flex]
HiDream E1
AI Judge Analysis
FLUX.2 [flex]
- + Successfully incorporates all requested elements: news anchor desk, hockey jerseys/stick, and multiple dogs.
- + Maintains the subject's brown hair color and clothing style from the source image.
- + Uses a clean, consistent 2D illustration style for the caricature.
- − The text on the news desk is slightly garbled ('THGIANAES NEWS').
HiDream E1
- + Features a 3D animated style that feels very 'exaggerated and humorous' as requested.
- + Good integration of the hockey theme with players in the background.
- − Changed the subject's hair from brown to blonde, losing resemblance to the source image.
- − There are noticeable artifacts on the hands, with extra fingers appearing on the dog's back.
- − The composition feels slightly cluttered around the subject's arms.
Verdict: FLUX.2 [flex] is the superior model because it better preserved the identity of the person in the source image, specifically their hair color and choice of casual denim clothing. While HiDream E1 provided a fun 3D style, it failed to maintain the subject's appearance and suffered from significant anatomical artifacts on the hands.
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 [flex]
HiDream E1
AI Judge Analysis
FLUX.2 [flex]
- + Perfectly captures the Studio Ghibli watercolor texture and soft, hand-painted aesthetic.
- + Preserves complex details from the original image like the plaid pattern on the shirt.
- + Maintains the correct facial expressions and emotional nuance of the source material.
- − The woman in pink has significantly more facial detail than characters typically found in Ghibli films.
HiDream E1
- + Adheres to a clean, digital illustration style.
- + Successfully translates the pose and composition of the original image.
- − Lacks the specific Ghibli hand-painted texture and soft pastel palette requested.
- − Loses significant detail, such as the plaid pattern on the man's shirt which is turned into a solid color.
- − The facial expression of the girlfriend (right) is changed from angry/disturbed to neutral/pleasant, losing the point of the meme.
Verdict: FLUX.2 [flex] (Model A) is the clear winner as it perfectly captures the specific 'hand-painted' and 'watercolor' style associated with Studio Ghibli, while preserving details like the plaid shirt and the original facial expressions. HiDream E1 (Model B) delivers a generic digital vector-style illustration that ignores the texture requests and fails to preserve the emotional context of the scene.
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 [flex]
HiDream E1
AI Judge Analysis
FLUX.2 [flex]
- + Excellent source preservation, maintaining the woman's face and the dog's appearance perfectly.
- + Realistic hair physics that look integrated into the scene.
- + The green leaves feel spatially consistent with the environment.
- − The leaves appear a bit static and lack motion blur to convey 'flying' speed.
HiDream E1
- + Successfully added hair motion and leaves as requested.
- + The red leaves provide a high-contrast pop of color.
- − Significantly altered the woman's facial features and the dog's face, failing to preserve the source identity.
- − The red leaves look like flat, superimposed stickers rather than objects in 3D space.
- − Loss of fine texture in the background and clothing compared to the source.
Verdict: FLUX.2 [flex] is the superior model for this task because it effectively applied the requested edits while keeping the original subjects identical to the source image. In contrast, HiDream E1 failed the primary requirement of image editing by noticeably changing the woman's face and the dog's appearance, while adding leaves that look poorly composited.
Next steps
Explore each model
HiDream AI's image-to-image editing model for instruction-based image modifications and transformations