Skip to content

Comment on How to generate realistic people in Stable Diffusionparent

Comments

RealvisXL is much better, probably the best model for photorealism. Example with same prompt: https://gencdn.aieasypic.com/original/ed7c2742-4a76-4e22-9bb...

It’s currently the best open weights model for prompt adherence too: https://imgsys.org/

It’s currently the best open weights model for prompt adherence too: https://imgsys.org/

This doesn't specifically measure prompt following but how good it is overall.

Is it also based on SD?

Pro tip, if you see a model with XL on the end, it's for SDXL, which is a stable diffusion model

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.