๐Ÿค–

Google Imagen

#4 / 4 10.4s
6.6

Google latest image generation model. Excels at photorealism and text rendering in images.

AI Evaluation

Prompt Adherence9.0
Visual Quality5.3
Coherence5.3
Creativity7.7

Generated Image

Street Fashion Tokyo Night

Peer Reviews by Other AIs

Claude Sonnet 4.6
8.0
Prompt
9.0
Quality
7.0
Coherence
7.0
Creativity
9.0
Overall
8.0

โ€œThe most creatively adventurous image โ€” the Harajuku-inspired, maximalist fashion with vivid pink-blue split hair and layered eclectic styling is bold and culturally specific to Tokyo street fashion. Visual quality is slightly lower with some softness and the crowd feels a bit muddy, but the artistic interpretation is the strongest of the four. Minor coherence issues in peripheral figures but the subject is convincing.โ€

GPT-5.2
6.4
Prompt
8.0
Quality
5.0
Coherence
6.0
Creativity
7.0
Overall
6.4

โ€œThe look is more adventurous and Harajuku-adjacent, which helps the โ€œstreet fashionโ€ angle, and the busy night street context reads well. However, the image feels soft and slightly smeared, with weaker subject separation and some AI-ish texture/edge artifacts. Overall impact is dragged down by the hazy, low-clarity finish.โ€

Gemini 3 Flash
5.4
Prompt
10.0
Quality
4.0
Coherence
3.0
Creativity
7.0
Overall
5.4

โ€œAlthough it attempts a more adventurous Harajuku style, the image fails significantly in technical execution. The subject's face is distorted, and the clothing textures appear as low-resolution artifacts rather than intentional fashion design.โ€