A Battle of Reflections: ChatGPT vs Google Gemini Nano Banana 2

Two powerful AI models tackle the same prompt: a pitch-black studio, broken mirrors, and infinite reflections. But which model delivers the most striking and accurate visual result? Dive into the comparisons of facial accuracy and emotional depth as each model interprets the scene with photorealistic detail and cinematic lighting.

Prompt

Use the facial identity given and keep the facial features exactly same, and generate a picture standing in a pitch-black studio, surrounded by dozens of broken mirror shards floating mid-air around him at different angles — each shard reflecting a different version of him, different expressions, different outfits, different emotions. He stands perfectly still in the center, wearing a crisp black turtleneck, arms at his sides, staring forward. The only light source is cold white light hitting each shard differently, fracturing the room into infinite reflections. Vertical 9:13 format. Ultra photorealistic, 8K, no text overlays, cinematic color grading.
Published: May 2, 2026 by