I used the same prompt and seed to generate images using 5 different checkpoints. Generally, the Generation settings are similar (DPM++ x, Karras, 30 steps, CFG Scale=5) except for the IntoRealism, which is designed differently (LCM, 8 steps, CFG=1). It takes about 3 minutes to render a batch of 2 pics on my PC.
Given the simple prompt here, all of these generate quite good images. The main difference seems to be lighting and shadows. I don't know how much real effect these LORAs have. LMK if you see any artifacts because I didn't notice any (beyond the text on the first one – forgot my negative prompts around watermarks and text).
The five checkpoints used (in order of the pics uploaded) were: bigLove_xl3, juggernautXL_ragnarokBy, pornmaster_proSDXLV5VAE, lustifySDXLNSFW, intorealismXL_v22
Prompt follows:
A sexy confident beautiful woman with thin limbs, skinny arms and legs, flat soft stomach, (wide hips). Large natural breasts, teardrop form, visible weight and gravity. Her figure is slim, tanned skin. Sultry expression, choker.
<lora:LabiaXXL:0.9>labiaxxl visible pussy lips <lora:longnipsxl:.9> (longnips)
She is squatting naked knees open, back arched, presenting, on a stage, evening glow lighting, cinematic bokeh, fantasy style, full body, full depth of field, head-to-toe visible, she is fully in focus.
