Do you still use older models?

Who here still uses older models, and what for? I still get a ton of use out of SD 1.4 and 1.5. They make great start images.

https://redd.it/1pkwms0
@rStableDiffusion

Z-Image: A bit of prompt engineering (prompt included)
https://redd.it/1pl6cxv
@rStableDiffusion

A “basics-only” guide to using ComfyUI the comfy way

https://redd.it/1pl6tlz
@rStableDiffusion

Chroma on itself kinda sux due to speed and image quality. Z-image kinda sux regarding artistic styles. both of them together kinda rules. small 768x1024 10 steps chroma image and 2 k zimage refiner.

https://redd.it/1plaaeo
@rStableDiffusion

What makes Z-image so good?

Im a bit of a noob when it comes to AI and image generation. Mostly watching different models generating images like qwen or sd.
I just use Nano banana for hobby.

Question i had was what makes Z-image so good? I know it can run efficiently on older gpus and generate good images but what prevents other models from doing the same.

tldr : what is Z-image doing differently?
Better training , better weights?

Question : what is the Z-image base what everyone is talking about? Next version of z-image



https://redd.it/1pldusz
@rStableDiffusion

Increase Your Level Of Details With Daemon Details Nodes and Generate Images at 4k With Z Img Turbo with DyPE
https://youtu.be/8rRCfhj4GBo

https://redd.it/1plh6dm
@rStableDiffusion