Wan 2.2 - Official Release!

A special thanks to tdrussell and and diffusion-pipe! They cracked the code. This version was trained with their newest settings for Wan 2.2. You need both the HIGH and LOW models (they are different). Run both with a strength of 1.0. All sample images were generated with the lightx2v lora. 12 steps, 6 on each model... 2 minutes to generate on an RTX 5090!!!

Workflow is in the 'training data' download file.

T2V Trigger: She is having an orgasm.

v2 fixes the issue with text overlays randomly popping up in the video... but I trained it at rank 64, like I have with some of my other LoRAs, but I feel like it is hindering this one a bit... So expect a v3 in the near future.

(Works for males too.)

T2V Trigger: She is having an orgasm.

I2V Trigger: She is having an orgasm. Her body shudders and shivers as a profound, guttural moan of sexual pleasure escapes her lips, her face a portrait of sublime ecstasy as she reaches the peak of her orgasm.

Example Prompt: A young White woman with fair skin and bright, blue eyes sits on a bench in a quiet, serene garden. She's wearing a pastel, floral dress with a lace trim, her blonde hair flowing in the gentle breeze. The camera pans over the colorful flowers and the gentle hum of bees before zooming in on her face. Her eyes close as she is having an orgasm, a soft, contented sigh escaping her lips.

abebes was kind enough to share their data set for their I2V orgasm LoRA. I expanded the data set, recaptioned it and trained it for T2V with my training workflow. Enjoy!

(Don't pay someone to train a LoRA for you. I can afford my own coffee. Send me your ideas and requests.)