⚠: 2s Is too short, Best Duration for the Lora to have full movement is 4s, even better with 6s.

Wantoe(Montoena) is a i2V Lora, allowing for easy Image to Video toe movement generation compatible with Wan 2.1 and Wan 2.2

All Videos are embedded with the same workflow.

The workflow is the Official Wan 2.2 i2v Comfyui workflow with no significant modification just the far better "Power lora loader" node and a "save image with metadata" node added.

Also work with Official Wan 2.1 i2v Comfyui workflow.

Lora Limitation and advice and best Practice :

⚠: The Lora Work only on front facing character.

Make sure that the feet in your original image is close enough to the foreground best distance can be show in most of the gallery examples. The further and the worst the result will be. Sometimes the generation will add a 6th toe or be disappointing, you have to re-roll.

This image also contain the workflow in case you can't read Video workflows⬇

Trigger : a girl flaunting her feet, WanT0e woman barefoot wiggling toes,

Other triggers (used to train the lora but work less well) can be :

barefoot girl toes curling

barefoot woman toe movement

WanT0e barefoot woman toe movement

WanT0e barefoot girl toes curling

WanT0e barefoot woman toe movement

WanT0e woman barefoot flexing toes

WanT0e woman barefoot wiggling toes