Uncensored Image-to-Video Tool Guide (WAN 2.1) for BJ and Cumshot [8GB VRAM][NVIDIA only]
Using Wan through Pinokio is by far the best way to use it if you do not want to spend hours trying to solve endless problems for essentially very similar results.
Install and launch Wan 2.1 image-to-video 480p
Step 1: Install Pinokio. Go toย https://pinokio.computer/
Step 2: Launch Pinokio. click the discover/browse icon and click wan 2.1 and click install in this window.
Step 3: click start after the installation is finished and select the wan2.1 image2video 480p model

Get Loras and start creating videos
For Image to blowjob fakes:
https://civitai.com/models/1358184/wan-21-penis-cock-dick-lora-t2v-i2v?modelVersionId=1534254
https://civitai.com/models/1333923/wan-pov-blowjob?modelVersionId=1547106
Add LoRAs to the I2V LoRA folder and refresh. (E:\pinokio\api\wan.git\app\loras_i2v)
Go to Loras at the bottom of the page and select loras:

Go to speed:

Set teacache to 2,5 x speed for faster generation.
Reference Example

Use prompt-based motion instructions:
As the scene begins, the woman pictured leans forward smoothly, while the camera gently pans downward, keeping her face centered in the frame. As she moves, a manโs pelvis with an erect penis attached is revealed, positioned steadily at the bottom center of the frame. The camera transitions into a POV angle, maintaining the male pelvis and penis consistently in the lower center of the scene. The focus remains on the woman’s facial expressions and movements. She immediately begins to perform a POV blowjob with fluid, deliberate motions, maintaining strong, playful eye contact with the camera. The camera stays tightly focused on her face while preserving the intimate, immersive perspective throughout.
What to expect
On 8GB Vram and 32 GB of Ram the proccess time for 5 seconds is 20 minutes averege, I don’t recommend going over 10 seconds with 8GB VRAM.
Expect slow renders on 8 GB VRAM; higher-end GPUs can handle longer clips and higher resolutions.