My own personal take on Wildering's artstyle. I feel it captured the mood and anatomy well enough for posting. More grounded and realistic feel overall.
I cannot generate locally for the time being, I'll be experimenting with Civitai's trainer and I'd be grateful for people to test it and post their results here.
Tested mainly on Nova Furry , following Crody's excellent Image Generation guide.

Model features:

• grounded and more realistic

• species anatomy

• cinematic, naturalistic rendering

• moderate visual contrast

Note: the LoRA performs best when the species is specified clearly in the prompt (wolf, dragon, shark, fox, etc.). Using only generic tags like anthro or furry leads to merged mammal/scalie traits or inconsistent creature features. So you would go anthro [sex] [species].

Dependant on your checkpoint, I use :
~0.6–0.8 strength, 25 steps, 15 hires with Remacri and CFG <=5

Weaknesses:

• can become over-stylized at high strengths

• can exaggerate anatomy and highlights, merging scalie/anthro features if pushed too hard

• may struggle with crowded scenes or highly complex interactions