Comparing Multiple Image Models Generating Sports Stars

gizakdag · x · 2026-07-19

This post showcases multiple images on the same theme comparing sports star images generated by Midjourney, Nano Banana Pro, GPT-2, Seedream 5 Pro, etc.

The key observation: differences in pose, style, cartoonishness, and detail fidelity across models, making it a display of multimodal generation capabilities.

Original post →

More from Multimodal

Multimodal channel →