


Reference poses will not update custom poses. Please scroll down to collect images. By default, you can choose 9 half-body images, 4 images, 8 images, 10 images, and 1 custom image, sitting pose 4 images.
LORA activation trigger words should be in English, specifically used to activate LORA. If you have modified the LORA, you can input the activation words here.
Add prompt words (in Chinese), for example, if you only uploaded a portrait. Then you can add prompt words yourself, such as long hair, white dress, flat shoes (note to turn off the weight of high heels), background description, etc.
Add chest adjustment (adjust between 5 to 5. Sometimes 5 may trigger platform restrictions. Try to use 2 or 3, 5 will make it completely flat.)
Add shiny silk (requires activating LORA words: shiny skin: Shiny skin, shiny or shiny pantyhose: Sparkling pantyhose, sparkling,
If white silk is not obvious, you can add prompt words ((white cotton leggings )),realistic, shiny skin, shiny, Shiny,
For black shiny silk, the prompt words can beShiny pantyhose, shiny, pantyhose, black pantyhose,),
Pointed red-bottom high heels element (activate LORA, prompt words or reference images containing high heels are sufficient),
Back view killer (requires prompt words to activate, it won't work without prompt words. Strength 1 to 1.4, keywords include Stand confidently with your back to us).
You can adjust or switch LORA yourself. The resulting images are guaranteed to be top-notch for gaining followers. Essential for platforms like Xiaohongshu, Douyin, and WeChat public accounts. Accelerated output: 14 minutes for 8 TTP HD images with facial repair. Non-members can generate 6 images within 10 minutes. When uploading custom pose images, make sure the images can be evenly divided. For example, a grid image that can be evenly divided into rows and columns. You can search for pose images on Baidu, then copy and upload them directly. For example, 2 rows with 5 images per row, you can select 2 rows and 5 columns to generate 10 images. If the image has 2 rows and 4 columns, select 2 rows and 4 columns in the custom row-column settings.
Select 1: 9 half-body images, Select 2: 4 full-body images, Select 3: 6 full-body images, Select 4: 8 full-body images,
Select 5: Default 1 full-body image, you can upload pose photos yourself,
Select 6: Sitting pose 4 reference images
Reference image is
Reference poses will not update custom poses. Please scroll down to collect images. By default, you can choose 9 half-body images, 4 images, 8 images, 10 images, and 1 custom image, sitting pose 4 images.
LORA activation trigger words should be in English, specifically used to activate LORA. If you have modified the LORA, you can input the activation words here.
Add prompt words (in Chinese), for example, if you only uploaded a portrait. Then you can add prompt words yourself, such as long hair, white dress, flat shoes (note to turn off the weight of high heels), background description, etc.
Add chest adjustment (adjust between 5 to 5. Sometimes 5 may trigger platform restrictions. Try to use 2 or 3, 5 will make it completely flat.)
Add shiny silk (requires activating LORA words: shiny skin: Shiny skin, shiny or shiny pantyhose: Sparkling pantyhose, sparkling,
If white silk is not obvious, you can add prompt words ((white cotton leggings )),realistic, shiny skin, shiny, Shiny,
For black shiny silk, the prompt words can beShiny pantyhose, shiny, pantyhose, black pantyhose,),
Pointed red-bottom high heels element (activate LORA, prompt words or reference images containing high heels are sufficient),
Back view killer (requires prompt words to activate, it won't work without prompt words. Strength 1 to 1.4, keywords include Stand confidently with your back to us).
You can adjust or switch LORA yourself. The resulting images are guaranteed to be top-notch for gaining followers. Essential for platforms like Xiaohongshu, Douyin, and WeChat public accounts. Accelerated output: 14 minutes for 8 TTP HD images with facial repair. Non-members can generate 6 images within 10 minutes. When uploading custom pose images, make sure the images can be evenly divided. For example, a grid image that can be evenly divided into rows and columns. You can search for pose images on Baidu, then copy and upload them directly. For example, 2 rows with 5 images per row, you can select 2 rows and 5 columns to generate 10 images. If the image has 2 rows and 4 columns, select 2 rows and 4 columns in the custom row-column settings.
Select 1: 9 half-body images, Select 2: 4 full-body images, Select 3: 6 full-body images, Select 4: 8 full-body images,
Select 5: Default 1 full-body image, you can upload pose photos yourself,
Select 6: Sitting pose 4 reference images
Reference image is