CivArchive
    WAN 2.2 S2V Lipsync Workflow with SageAttention + BlockSwap + GGUF (include Upscale and Interpolation) - v2.0 (GGUF)
    NSFW

    My TG Channel - https://t.me/StefanFalkokAI

    My TG Chat - https://t.me/+y4R5JybDZcFjMjFi

    Hi! I introduce my working workflow with Wan 2.2 S2V generation video for ComfyUI.

    You need to have Wan 2.2 S2V model

    You need to have Wan 2.2 models (https://huggingface.co/Kijai/WanVideo_comfy_fp8_scaled/tree/main/S2V), clip (https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/main/split_files/text_encoders/umt5_xxl_fp16.safetensors), audio_encoder (https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/tree/main/split_files/audio_encoders) and vae (wan 2.1 vae)

    GGUF - https://huggingface.co/QuantStack/Wan2.2-S2V-14B-GGUF/tree/main

    Also I have included Lora https://huggingface.co/Kijai/WanVideo_comfy/tree/main/Lightx2v for better and fast generation

    I recomend set strength 1.5 for loras to get best results

    download my ComfyUI Build https://huggingface.co/datasets/StefanFalkok/ComfyUI_portable_torch_2.9.1_cu130_cp313_sageattention_triton, also u need download and install CUDA 13.0 (https://developer.nvidia.com/cuda-13-0-0-download-archive) and VS Code (https://visualstudio.microsoft.com/downloads/)

    just use audio, image, write prompt and enjoy!

    Leave comments if you have trouble or you found the problem with workflow

    Description

    FAQ

    Comments (4)

    gmlwn11795Oct 18, 2025
    CivitAI

    KSamplerAdvanced

    mat1 and mat2 shapes cannot be multiplied (154x768 and 4096x5120)

    Stefan_Falkok
    Author
    Oct 18, 2025

    Change clip to umt5_xxl_enc_bf16, don't use fp8 version

    gmlwn11795Oct 18, 2025

    @Stefan_Falkok I got an error when using the umt5_xxl_enc_bf16.safetensors version linked on Civitai.
    However, when I switched to umt5_xxl_fp16.safetensors from the workflow’s own link, the error disappeared.
    All other models were loaded exactly from the same links provided in the workflow.

    Additionally, the generated video starts with a bright flash or color distortion on the very first frame
    is there any way to fix or reduce that issue?

    Stefan_Falkok
    Author
    Oct 18, 2025

    @gmlwn11795 i can't find solution of bright flash frame - it's a sore of the model with first frame

    Workflows
    Wan Video 2.2 I2V-A14B

    Details

    Downloads
    578
    Platform
    CivitAI
    Platform Status
    Available
    Created
    10/17/2025
    Updated
    4/28/2026
    Deleted
    -

    Files

    wan22S2VLipsyncWorkflowWith_v20GGUF.zip

    Mirrors