FL4X Alpha!

Welcome to FL4X alpha! This model aims to offer higher contrast, vibrant colors, and better prompt understanding. It was the result of merges and additional training. Roughly ~30 hours of GPU time were used to tweak/fine-tune this model. I am moving some of my training to Civitai shortly.

Comparisons

Recommended Settings

I recommend using the KSampler node in ComfyUI. This workflow works well: