GPT Image 1 OpenAI HiDream E1 HiDream AI

Settled by community votes across 4 shared challenges, with an AI judge weighing in on each.

GPT Image 1

22.9 arena score

#29 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

GPT Image 1

win rate

Ties

HiDream E1

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 Editing

Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

GPT Image 1

HiDream E1

AI Judge Analysis

GPT Image 1

+ Excellent source preservation, keeping the original face, clothes, and background intact.
+ The hair texture and color match the existing beard perfectly.
+ The hairline follows the natural anatomy of the forehead well.

− The hair volume is slightly tall, giving it a somewhat 'wig-like' silhouette at the very top.

HiDream E1

+ Successfully added a large volume of hair as requested.

− Significantly altered the background, changing the landscape to a desert with mountains.
− The hair texture appears synthetic and overly smooth compared to the rugged source image.
− Heavy artifacts and 'deep-fried' texture added to the skin and jacket.

Verdict: GPT Image 1 is the clear winner as it successfully performs the edit while maintaining perfect source preservation of the person's face and the background. HiDream E1 fails the editing task by completely regenerating the environment and introducing harsh digital artifacts to the subject's skin and clothing.

Over-the-top cartoon caricature

Editing

Edit 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

GPT Image 1

HiDream E1

AI Judge Analysis

GPT Image 1

+ Excellent preservation of the subject's features and hair color in caricature form
+ Authentic caricature art style with humorous exaggeration
+ Cleverly integrates all three elements: news desk, dog companion, and hockey imagery on screen.

− The hand holding the paper looks slightly unfinished compared to the rest of the drawing.

HiDream E1

+ Successfully incorporates the dog and hockey rink background
+ High resolution and clean digital rendering style

− Fails to preserve the likeness of the original woman, significantly changing her hair color to blonde
− The 'caricature' style leans more towards a generic 3D animation character rather than a specific caricature of the subject
− Missing the 'TV show anchor' profession element beyond a generic blazer.

Verdict: GPT Image 1 is the clear winner as it successfully maintains the subject's identity while transforming her into a hand-drawn caricature. It cleverly integrates her profession as a news anchor and her hobbies, whereas HiDream E1 fails to preserve the source image's features and ignores the specific instruction to show her as a TV anchor.

Studio Ghibli Anime Style

Editing

Edit 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

GPT Image 1

HiDream E1

AI Judge Analysis

GPT Image 1

+ Excellent adherence to the 'hand-painted textures' and 'soft pastel' request.
+ Accurately recreates the specific facial expressions and lighting of the source image.
+ Preserves original clothing colors and patterns (like the blue plaid shirt) very well.

− The textures are a bit grainy, leaning more towards colored pencil or watercolor than clean cel-shading.

HiDream E1

+ Features very clean line art and a clear anime aesthetic.
+ Successfully captures the Ghibli-esque rounded character designs.

− Fails to preserve original colors, changing the iconic red dress to peach and the blue shirt to pink.
− The lighting is flat and lacks the 'warm, nostalgic mood' requested in the prompt.
− Characters have generic expressions that don't match the specific emotions in the source photo.

Verdict: GPT Image 1 is the superior edit because it successfully translates the specific details and emotions of the 'distracted boyfriend' meme into the requested style while preserving identifying features like the plaid shirt and red dress. HiDream E1 provides a generic anime filter that ignores the original color palette and the specific facial expressions that make the source image recognizable.

Golden Hour Stroll

Image Editing

Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

GPT Image 1

HiDream E1

AI Judge Analysis

GPT Image 1

+ Excellent preservation of the subject's identity and facial features
+ Motion blur on the leaves adds to the dynamic feel
+ Realistic hair movement that blends naturally with the original hair texture

− The leaf debris is quite small and looks more like mulch than falling autumn leaves
− The background bridge area becomes slightly blurred or distorted compared to the source

HiDream E1

+ Very clear and energetic depiction of blowing hair
+ Large, vibrant leaves make the edit immediately obvious

− Significant loss of source preservation; the face and dog look noticeably different from the original image
− The red leaves look like flat, artificial stickers layered on top of the image
− Lighting on the subject is overly saturated and loses the natural feel of the source

Verdict: GPT Image 1 is the superior edit because it maintains the integrity of the original photo while successfully integrating the requested motion effects. HiDream E1 fails as an editing model by significantly altering the woman's face and the dog's appearance, and its added elements lack the realistic integration seen in GPT Image 1.