# I need some real advice on how to improve my realistic and natural AI — retained record

> Retained reddit record from reddit:StableDiffusion.

Canonical: https://highsignal.app/data/records/1c611a1776d56d6a

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:StableDiffusion)  published 2026-10-09 11:30 ingested 2026-10-10 02:36

[original source ↗](https://www.reddit.com/r/StableDiffusion/comments/1x1ixlv/i_need_some_real_advice_on_how_to_improve_my/)

## retained text



```
Hello, 

I'll keep it short and simple : 

My goal is to create a natural, realistic result like a selfie, with natural lighting, posing, and surroundings, that doesn't look like plastic AI or seem a bit slapdash. Everyday. 

I’m currently using Krea 2 Turbo with two LORAs. (I work on Forge NEO), I like the results, but I want to know what you recommend to achieve this look and whether I’m doing it right.   
Because I'm seeing some amazing results here ! Is upscaling necessary ? Adetailers ? Double sampling ? 

And another question: I know Krea 2 is newer and better than Z Image Turbo. But is Krea 2 Turbo still the best option right now ? Or is Qwen 2.1 better for this ? 

AI is evolving very quickly are there any websites where I can keep track of its progress and new releases? Like a new, realistic, and natural local AI model coming out, a LORA, etc.
```



document fetched 2026-10-10 00:32 · canonical https://reddit.com/r/StableDiffusion/comments/1x1ixlv/i_need_some_real_advice_on_how_to_improve_my
