Loras AI
Speed

LCM-LoRA SDXL LoRA for SDXL

An LCM-LoRA adapter that brings latent consistency distillation to SDXL, cutting generation to roughly 2–8 steps.

Details

Base model
SDXL
Type
Speed
Author
latent-consistency
License
OpenRAIL++
Trigger words
None needed

How to use it

Load it on SDXL 1.0 base with the LCM scheduler, and use a low step count and low guidance as described on the model card.

License

Released under an OpenRAIL license: commercial use is allowed, but its use restrictions (no unlawful, harmful or deceptive use) apply to you and to anyone you share it with. Check the base model’s own license too.

Questions about this LoRA

How do I use LCM-LoRA SDXL?

Load it on SDXL 1.0 base with the LCM scheduler, and use a low step count and low guidance as described on the model card.

Can I use LCM-LoRA SDXL commercially?

Yes. LCM-LoRA SDXL is released under OpenRAIL++, which allows commercial use subject to its use restrictions, such as no unlawful or harmful use. Credit latent-consistency and also check the license of SDXL.