← All benchmarks

Tokyo Alley After Rain

Three AI models render a rain-soaked Shinjuku alley with stacked neon signs reflecting in puddles. Stress-tests wet-surface reflections + colour-light interaction.

7 models compared · same input · same prompt

Input & prompt

Prompt template

A narrow Tokyo back-alley late at night, just after heavy rain. Looking down the alley's length. The model populates the scene — what kind of signs are on the walls, what the puddles reflect, what (if anything) inhabits the frame. Low-light photography. Wet surfaces, not dry. Night, not day.

ChatGPT (GPT Image 2)

Try this →
76.1sopenai/gpt-image-2

Nano Banana 2 (Gemini)

Try this →
15.0sgoogle/nano-banana-2

Grok Imagine Pro

Try this →
13.1sx-ai/grok-imagine-image-pro

FLUX.2 Max

Try this →

Flux Kontext Pro

Try this →

Seedream 5 Lite

Try this →

Nano Banana 2 Lite

Try this →
google/gemini-3.1-flash-lite-image

Want to try this with your own twist?

Run the same prompt with your own variables — it's free.

Try the prompt →